feat: connect to spicedb

106 files changed, 18195 insertions, 0 deletions

diff --git a/vendor/github.com/charmbracelet/colorprofile/.golangci-soft.yml b/vendor/github.com/charmbracelet/colorprofile/.golangci-soft.yml
new file mode 100644
index 0000000..d325d4f
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/.golangci-soft.yml
@@ -0,0 +1,40 @@
+run:
+  tests: false
+  issues-exit-code: 0
+
+issues:
+  include:
+    - EXC0001
+    - EXC0005
+    - EXC0011
+    - EXC0012
+    - EXC0013
+
+  max-issues-per-linter: 0
+  max-same-issues: 0
+
+linters:
+  enable:
+    - exhaustive
+    - goconst
+    - godot
+    - godox
+    - mnd
+    - gomoddirectives
+    - goprintffuncname
+    - misspell
+    - nakedret
+    - nestif
+    - noctx
+    - nolintlint
+    - prealloc
+    - wrapcheck
+
+  # disable default linters, they are already enabled in .golangci.yml
+  disable:
+    - errcheck
+    - gosimple
+    - govet
+    - ineffassign
+    - staticcheck
+    - unused
diff --git a/vendor/github.com/charmbracelet/colorprofile/.golangci.yml b/vendor/github.com/charmbracelet/colorprofile/.golangci.yml
new file mode 100644
index 0000000..d6789e0
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/.golangci.yml
@@ -0,0 +1,28 @@
+run:
+  tests: false
+
+issues:
+  include:
+    - EXC0001
+    - EXC0005
+    - EXC0011
+    - EXC0012
+    - EXC0013
+
+  max-issues-per-linter: 0
+  max-same-issues: 0
+
+linters:
+  enable:
+    - bodyclose
+    - gofumpt
+    - goimports
+    - gosec
+    - nilerr
+    - revive
+    - rowserrcheck
+    - sqlclosecheck
+    - tparallel
+    - unconvert
+    - unparam
+    - whitespace
diff --git a/vendor/github.com/charmbracelet/colorprofile/.goreleaser.yml b/vendor/github.com/charmbracelet/colorprofile/.goreleaser.yml
new file mode 100644
index 0000000..40d9f29
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/.goreleaser.yml
@@ -0,0 +1,6 @@
+includes:
+  - from_url:
+      url: charmbracelet/meta/main/goreleaser-lib.yaml
+
+# yaml-language-server: $schema=https://goreleaser.com/static/schema-pro.json
+
diff --git a/vendor/github.com/charmbracelet/colorprofile/LICENSE b/vendor/github.com/charmbracelet/colorprofile/LICENSE
new file mode 100644
index 0000000..b7974b0
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2020-2024 Charmbracelet, Inc
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/vendor/github.com/charmbracelet/colorprofile/README.md b/vendor/github.com/charmbracelet/colorprofile/README.md
new file mode 100644
index 0000000..c72b2f4
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/README.md
@@ -0,0 +1,103 @@
+# Colorprofile
+
+<p>
+    <a href="https://github.com/charmbracelet/colorprofile/releases"><img src="https://img.shields.io/github/release/charmbracelet/colorprofile.svg" alt="Latest Release"></a>
+    <a href="https://pkg.go.dev/github.com/charmbracelet/colorprofile?tab=doc"><img src="https://godoc.org/github.com/charmbracelet/colorprofile?status.svg" alt="GoDoc"></a>
+    <a href="https://github.com/charmbracelet/colorprofile/actions"><img src="https://github.com/charmbracelet/colorprofile/actions/workflows/build.yml/badge.svg" alt="Build Status"></a>
+</p>
+
+A simple, powerful—and at times magical—package for detecting terminal color
+profiles and performing color (and CSI) degradation.
+
+## Detecting the terminal’s color profile
+
+Detecting the terminal’s color profile is easy.
+
+```go
+import "github.com/charmbracelet/colorprofile"
+
+// Detect the color profile. If you’re planning on writing to stderr you'd want
+// to use os.Stderr instead.
+p := colorprofile.Detect(os.Stdout, os.Environ())
+
+// Comment on the profile.
+fmt.Printf("You know, your colors are quite %s.", func() string {
+    switch p {
+    case colorprofile.TrueColor:
+        return "fancy"
+    case colorprofile.ANSI256:
+        return "1990s fancy"
+    case colorprofile.ANSI:
+        return "normcore"
+    case colorprofile.Ascii:
+        return "ancient"
+    case colorprofile.NoTTY:
+        return "naughty!"
+    }
+    return "...IDK" // this should never happen
+}())
+```
+
+## Downsampling colors
+
+When necessary, colors can be downsampled to a given profile, or manually
+downsampled to a specific profile.
+
+```go
+p := colorprofile.Detect(os.Stdout, os.Environ())
+c := color.RGBA{0x6b, 0x50, 0xff, 0xff} // #6b50ff
+
+// Downsample to the detected profile, when necessary.
+convertedColor := p.Convert(c)
+
+// Or manually convert to a given profile.
+ansi256Color := colorprofile.ANSI256.Convert(c)
+ansiColor := colorprofile.ANSI.Convert(c)
+noColor := colorprofile.Ascii.Convert(c)
+noANSI := colorprofile.NoTTY.Convert(c)
+```
+
+## Automatic downsampling with a Writer
+
+You can also magically downsample colors in ANSI output, when necessary. If
+output is not a TTY ANSI will be dropped entirely.
+
+```go
+myFancyANSI := "\x1b[38;2;107;80;255mCute \x1b[1;3mpuppy!!\x1b[m"
+
+// Automatically downsample for the terminal at stdout.
+w := colorprofile.NewWriter(os.Stdout, os.Environ())
+fmt.Fprintf(w, myFancyANSI)
+
+// Downsample to 4-bit ANSI.
+w.Profile = colorprofile.ANSI
+fmt.Fprintf(w, myFancyANSI)
+
+// Ascii-fy, no colors.
+w.Profile = colorprofile.Ascii
+fmt.Fprintf(w, myFancyANSI)
+
+// Strip ANSI altogether.
+w.Profile = colorprofile.NoTTY
+fmt.Fprintf(w, myFancyANSI) // not as fancy
+```
+
+## Feedback
+
+We’d love to hear your thoughts on this project. Feel free to drop us a note!
+
+- [Twitter](https://twitter.com/charmcli)
+- [The Fediverse](https://mastodon.social/@charmcli)
+- [Discord](https://charm.sh/chat)
+
+## License
+
+[MIT](https://github.com/charmbracelet/bubbletea/raw/master/LICENSE)
+
+---
+
+Part of [Charm](https://charm.sh).
+
+<a href="https://charm.sh/"><img alt="The Charm logo" src="https://stuff.charm.sh/charm-badge.jpg" width="400"></a>
+
+Charm热爱开源 • Charm loves open source • نحنُ نحب المصادر المفتوحة
diff --git a/vendor/github.com/charmbracelet/colorprofile/env.go b/vendor/github.com/charmbracelet/colorprofile/env.go
new file mode 100644
index 0000000..8df3d8f
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/env.go
@@ -0,0 +1,287 @@
+package colorprofile
+
+import (
+	"bytes"
+	"io"
+	"os/exec"
+	"runtime"
+	"strconv"
+	"strings"
+
+	"github.com/charmbracelet/x/term"
+	"github.com/xo/terminfo"
+)
+
+// Detect returns the color profile based on the terminal output, and
+// environment variables. This respects NO_COLOR, CLICOLOR, and CLICOLOR_FORCE
+// environment variables.
+//
+// The rules as follows:
+//   - TERM=dumb is always treated as NoTTY unless CLICOLOR_FORCE=1 is set.
+//   - If COLORTERM=truecolor, and the profile is not NoTTY, it gest upgraded to TrueColor.
+//   - Using any 256 color terminal (e.g. TERM=xterm-256color) will set the profile to ANSI256.
+//   - Using any color terminal (e.g. TERM=xterm-color) will set the profile to ANSI.
+//   - Using CLICOLOR=1 without TERM defined should be treated as ANSI if the
+//     output is a terminal.
+//   - NO_COLOR takes precedence over CLICOLOR/CLICOLOR_FORCE, and will disable
+//     colors but not text decoration, i.e. bold, italic, faint, etc.
+//
+// See https://no-color.org/ and https://bixense.com/clicolors/ for more information.
+func Detect(output io.Writer, env []string) Profile {
+	out, ok := output.(term.File)
+	isatty := ok && term.IsTerminal(out.Fd())
+	environ := newEnviron(env)
+	term := environ.get("TERM")
+	isDumb := term == "dumb"
+	envp := colorProfile(isatty, environ)
+	if envp == TrueColor || envNoColor(environ) {
+		// We already know we have TrueColor, or NO_COLOR is set.
+		return envp
+	}
+
+	if isatty && !isDumb {
+		tip := Terminfo(term)
+		tmuxp := tmux(environ)
+
+		// Color profile is the maximum of env, terminfo, and tmux.
+		return max(envp, max(tip, tmuxp))
+	}
+
+	return envp
+}
+
+// Env returns the color profile based on the terminal environment variables.
+// This respects NO_COLOR, CLICOLOR, and CLICOLOR_FORCE environment variables.
+//
+// The rules as follows:
+//   - TERM=dumb is always treated as NoTTY unless CLICOLOR_FORCE=1 is set.
+//   - If COLORTERM=truecolor, and the profile is not NoTTY, it gest upgraded to TrueColor.
+//   - Using any 256 color terminal (e.g. TERM=xterm-256color) will set the profile to ANSI256.
+//   - Using any color terminal (e.g. TERM=xterm-color) will set the profile to ANSI.
+//   - Using CLICOLOR=1 without TERM defined should be treated as ANSI if the
+//     output is a terminal.
+//   - NO_COLOR takes precedence over CLICOLOR/CLICOLOR_FORCE, and will disable
+//     colors but not text decoration, i.e. bold, italic, faint, etc.
+//
+// See https://no-color.org/ and https://bixense.com/clicolors/ for more information.
+func Env(env []string) (p Profile) {
+	return colorProfile(true, newEnviron(env))
+}
+
+func colorProfile(isatty bool, env environ) (p Profile) {
+	isDumb := env.get("TERM") == "dumb"
+	envp := envColorProfile(env)
+	if !isatty || isDumb {
+		// Check if the output is a terminal.
+		// Treat dumb terminals as NoTTY
+		p = NoTTY
+	} else {
+		p = envp
+	}
+
+	if envNoColor(env) && isatty {
+		if p > Ascii {
+			p = Ascii
+		}
+		return
+	}
+
+	if cliColorForced(env) {
+		if p < ANSI {
+			p = ANSI
+		}
+		if envp > p {
+			p = envp
+		}
+
+		return
+	}
+
+	if cliColor(env) {
+		if isatty && !isDumb && p < ANSI {
+			p = ANSI
+		}
+	}
+
+	return p
+}
+
+// envNoColor returns true if the environment variables explicitly disable color output
+// by setting NO_COLOR (https://no-color.org/).
+func envNoColor(env environ) bool {
+	noColor, _ := strconv.ParseBool(env.get("NO_COLOR"))
+	return noColor
+}
+
+func cliColor(env environ) bool {
+	cliColor, _ := strconv.ParseBool(env.get("CLICOLOR"))
+	return cliColor
+}
+
+func cliColorForced(env environ) bool {
+	cliColorForce, _ := strconv.ParseBool(env.get("CLICOLOR_FORCE"))
+	return cliColorForce
+}
+
+func colorTerm(env environ) bool {
+	colorTerm := strings.ToLower(env.get("COLORTERM"))
+	return colorTerm == "truecolor" || colorTerm == "24bit" ||
+		colorTerm == "yes" || colorTerm == "true"
+}
+
+// envColorProfile returns infers the color profile from the environment.
+func envColorProfile(env environ) (p Profile) {
+	term, ok := env.lookup("TERM")
+	if !ok || len(term) == 0 || term == "dumb" {
+		p = NoTTY
+		if runtime.GOOS == "windows" {
+			// Use Windows API to detect color profile. Windows Terminal and
+			// cmd.exe don't define $TERM.
+			if wcp, ok := windowsColorProfile(env); ok {
+				p = wcp
+			}
+		}
+	} else {
+		p = ANSI
+	}
+
+	parts := strings.Split(term, "-")
+	switch parts[0] {
+	case "alacritty",
+		"contour",
+		"foot",
+		"ghostty",
+		"kitty",
+		"rio",
+		"st",
+		"wezterm":
+		return TrueColor
+	case "xterm":
+		if len(parts) > 1 {
+			switch parts[1] {
+			case "ghostty", "kitty":
+				// These terminals can be defined as xterm-TERMNAME
+				return TrueColor
+			}
+		}
+	case "tmux", "screen":
+		if p < ANSI256 {
+			p = ANSI256
+		}
+	}
+
+	if isCloudShell, _ := strconv.ParseBool(env.get("GOOGLE_CLOUD_SHELL")); isCloudShell {
+		return TrueColor
+	}
+
+	// GNU Screen doesn't support TrueColor
+	// Tmux doesn't support $COLORTERM
+	if colorTerm(env) && !strings.HasPrefix(term, "screen") && !strings.HasPrefix(term, "tmux") {
+		return TrueColor
+	}
+
+	if strings.HasSuffix(term, "256color") && p < ANSI256 {
+		p = ANSI256
+	}
+
+	return
+}
+
+// Terminfo returns the color profile based on the terminal's terminfo
+// database. This relies on the Tc and RGB capabilities to determine if the
+// terminal supports TrueColor.
+// If term is empty or "dumb", it returns NoTTY.
+func Terminfo(term string) (p Profile) {
+	if len(term) == 0 || term == "dumb" {
+		return NoTTY
+	}
+
+	p = ANSI
+	ti, err := terminfo.Load(term)
+	if err != nil {
+		return
+	}
+
+	extbools := ti.ExtBoolCapsShort()
+	if _, ok := extbools["Tc"]; ok {
+		return TrueColor
+	}
+
+	if _, ok := extbools["RGB"]; ok {
+		return TrueColor
+	}
+
+	return
+}
+
+// Tmux returns the color profile based on `tmux info` output. Tmux supports
+// overriding the terminal's color capabilities, so this function will return
+// the color profile based on the tmux configuration.
+func Tmux(env []string) Profile {
+	return tmux(newEnviron(env))
+}
+
+// tmux returns the color profile based on the tmux environment variables.
+func tmux(env environ) (p Profile) {
+	if tmux, ok := env.lookup("TMUX"); !ok || len(tmux) == 0 {
+		// Not in tmux
+		return NoTTY
+	}
+
+	// Check if tmux has either Tc or RGB capabilities. Otherwise, return
+	// ANSI256.
+	p = ANSI256
+	cmd := exec.Command("tmux", "info")
+	out, err := cmd.Output()
+	if err != nil {
+		return
+	}
+
+	for _, line := range bytes.Split(out, []byte("\n")) {
+		if (bytes.Contains(line, []byte("Tc")) || bytes.Contains(line, []byte("RGB"))) &&
+			bytes.Contains(line, []byte("true")) {
+			return TrueColor
+		}
+	}
+
+	return
+}
+
+// environ is a map of environment variables.
+type environ map[string]string
+
+// newEnviron returns a new environment map from a slice of environment
+// variables.
+func newEnviron(environ []string) environ {
+	m := make(map[string]string, len(environ))
+	for _, e := range environ {
+		parts := strings.SplitN(e, "=", 2)
+		var value string
+		if len(parts) == 2 {
+			value = parts[1]
+		}
+		m[parts[0]] = value
+	}
+	return m
+}
+
+// lookup returns the value of an environment variable and a boolean indicating
+// if it exists.
+func (e environ) lookup(key string) (string, bool) {
+	v, ok := e[key]
+	return v, ok
+}
+
+// get returns the value of an environment variable and empty string if it
+// doesn't exist.
+func (e environ) get(key string) string {
+	v, _ := e.lookup(key)
+	return v
+}
+
+func max[T ~byte | ~int](a, b T) T {
+	if a > b {
+		return a
+	}
+	return b
+}
diff --git a/vendor/github.com/charmbracelet/colorprofile/env_other.go b/vendor/github.com/charmbracelet/colorprofile/env_other.go
new file mode 100644
index 0000000..080994b
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/env_other.go
@@ -0,0 +1,8 @@
+//go:build !windows
+// +build !windows
+
+package colorprofile
+
+func windowsColorProfile(map[string]string) (Profile, bool) {
+	return 0, false
+}
diff --git a/vendor/github.com/charmbracelet/colorprofile/env_windows.go b/vendor/github.com/charmbracelet/colorprofile/env_windows.go
new file mode 100644
index 0000000..3b9c28f
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/env_windows.go
@@ -0,0 +1,45 @@
+//go:build windows
+// +build windows
+
+package colorprofile
+
+import (
+	"strconv"
+
+	"golang.org/x/sys/windows"
+)
+
+func windowsColorProfile(env map[string]string) (Profile, bool) {
+	if env["ConEmuANSI"] == "ON" {
+		return TrueColor, true
+	}
+
+	if len(env["WT_SESSION"]) > 0 {
+		// Windows Terminal supports TrueColor
+		return TrueColor, true
+	}
+
+	major, _, build := windows.RtlGetNtVersionNumbers()
+	if build < 10586 || major < 10 {
+		// No ANSI support before WindowsNT 10 build 10586
+		if len(env["ANSICON"]) > 0 {
+			ansiconVer := env["ANSICON_VER"]
+			cv, err := strconv.Atoi(ansiconVer)
+			if err != nil || cv < 181 {
+				// No 8 bit color support before ANSICON 1.81
+				return ANSI, true
+			}
+
+			return ANSI256, true
+		}
+
+		return NoTTY, true
+	}
+
+	if build < 14931 {
+		// No true color support before build 14931
+		return ANSI256, true
+	}
+
+	return TrueColor, true
+}
diff --git a/vendor/github.com/charmbracelet/colorprofile/profile.go b/vendor/github.com/charmbracelet/colorprofile/profile.go
new file mode 100644
index 0000000..97e37ac
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/profile.go
@@ -0,0 +1,399 @@
+package colorprofile
+
+import (
+	"image/color"
+	"math"
+
+	"github.com/charmbracelet/x/ansi"
+	"github.com/lucasb-eyer/go-colorful"
+)
+
+// Profile is a color profile: NoTTY, Ascii, ANSI, ANSI256, or TrueColor.
+type Profile byte
+
+const (
+	// NoTTY, not a terminal profile.
+	NoTTY Profile = iota
+	// Ascii, uncolored profile.
+	Ascii //nolint:revive
+	// ANSI, 4-bit color profile.
+	ANSI
+	// ANSI256, 8-bit color profile.
+	ANSI256
+	// TrueColor, 24-bit color profile.
+	TrueColor
+)
+
+// String returns the string representation of a Profile.
+func (p Profile) String() string {
+	switch p {
+	case TrueColor:
+		return "TrueColor"
+	case ANSI256:
+		return "ANSI256"
+	case ANSI:
+		return "ANSI"
+	case Ascii:
+		return "Ascii"
+	case NoTTY:
+		return "NoTTY"
+	}
+	return "Unknown"
+}
+
+// Convert transforms a given Color to a Color supported within the Profile.
+func (p Profile) Convert(c color.Color) color.Color {
+	if p <= Ascii {
+		return nil
+	}
+
+	switch c := c.(type) {
+	case ansi.BasicColor:
+		return c
+
+	case ansi.ExtendedColor:
+		if p == ANSI {
+			return ansi256ToANSIColor(c)
+		}
+		return c
+
+	case ansi.TrueColor, color.Color:
+		h, ok := colorful.MakeColor(c)
+		if !ok {
+			return nil
+		}
+		if p != TrueColor {
+			ac := hexToANSI256Color(h)
+			if p == ANSI {
+				return ansi256ToANSIColor(ac)
+			}
+			return ac
+		}
+		return c
+	}
+
+	return c
+}
+
+func hexToANSI256Color(c colorful.Color) ansi.ExtendedColor {
+	v2ci := func(v float64) int {
+		if v < 48 {
+			return 0
+		}
+		if v < 115 {
+			return 1
+		}
+		return int((v - 35) / 40)
+	}
+
+	// Calculate the nearest 0-based color index at 16..231
+	r := v2ci(c.R * 255.0) // 0..5 each
+	g := v2ci(c.G * 255.0)
+	b := v2ci(c.B * 255.0)
+	ci := 36*r + 6*g + b /* 0..215 */
+
+	// Calculate the represented colors back from the index
+	i2cv := [6]int{0, 0x5f, 0x87, 0xaf, 0xd7, 0xff}
+	cr := i2cv[r] // r/g/b, 0..255 each
+	cg := i2cv[g]
+	cb := i2cv[b]
+
+	// Calculate the nearest 0-based gray index at 232..255
+	var grayIdx int
+	average := (cr + cg + cb) / 3
+	if average > 238 {
+		grayIdx = 23
+	} else {
+		grayIdx = (average - 3) / 10 // 0..23
+	}
+	gv := 8 + 10*grayIdx // same value for r/g/b, 0..255
+
+	// Return the one which is nearer to the original input rgb value
+	c2 := colorful.Color{R: float64(cr) / 255.0, G: float64(cg) / 255.0, B: float64(cb) / 255.0}
+	g2 := colorful.Color{R: float64(gv) / 255.0, G: float64(gv) / 255.0, B: float64(gv) / 255.0}
+	colorDist := c.DistanceHSLuv(c2)
+	grayDist := c.DistanceHSLuv(g2)
+
+	if colorDist <= grayDist {
+		return ansi.ExtendedColor(16 + ci) //nolint:gosec
+	}
+	return ansi.ExtendedColor(232 + grayIdx) //nolint:gosec
+}
+
+func ansi256ToANSIColor(c ansi.ExtendedColor) ansi.BasicColor {
+	var r int
+	md := math.MaxFloat64
+
+	h, _ := colorful.Hex(ansiHex[c])
+	for i := 0; i <= 15; i++ {
+		hb, _ := colorful.Hex(ansiHex[i])
+		d := h.DistanceHSLuv(hb)
+
+		if d < md {
+			md = d
+			r = i
+		}
+	}
+
+	return ansi.BasicColor(r) //nolint:gosec
+}
+
+// RGB values of ANSI colors (0-255).
+var ansiHex = []string{
+	"#000000",
+	"#800000",
+	"#008000",
+	"#808000",
+	"#000080",
+	"#800080",
+	"#008080",
+	"#c0c0c0",
+	"#808080",
+	"#ff0000",
+	"#00ff00",
+	"#ffff00",
+	"#0000ff",
+	"#ff00ff",
+	"#00ffff",
+	"#ffffff",
+	"#000000",
+	"#00005f",
+	"#000087",
+	"#0000af",
+	"#0000d7",
+	"#0000ff",
+	"#005f00",
+	"#005f5f",
+	"#005f87",
+	"#005faf",
+	"#005fd7",
+	"#005fff",
+	"#008700",
+	"#00875f",
+	"#008787",
+	"#0087af",
+	"#0087d7",
+	"#0087ff",
+	"#00af00",
+	"#00af5f",
+	"#00af87",
+	"#00afaf",
+	"#00afd7",
+	"#00afff",
+	"#00d700",
+	"#00d75f",
+	"#00d787",
+	"#00d7af",
+	"#00d7d7",
+	"#00d7ff",
+	"#00ff00",
+	"#00ff5f",
+	"#00ff87",
+	"#00ffaf",
+	"#00ffd7",
+	"#00ffff",
+	"#5f0000",
+	"#5f005f",
+	"#5f0087",
+	"#5f00af",
+	"#5f00d7",
+	"#5f00ff",
+	"#5f5f00",
+	"#5f5f5f",
+	"#5f5f87",
+	"#5f5faf",
+	"#5f5fd7",
+	"#5f5fff",
+	"#5f8700",
+	"#5f875f",
+	"#5f8787",
+	"#5f87af",
+	"#5f87d7",
+	"#5f87ff",
+	"#5faf00",
+	"#5faf5f",
+	"#5faf87",
+	"#5fafaf",
+	"#5fafd7",
+	"#5fafff",
+	"#5fd700",
+	"#5fd75f",
+	"#5fd787",
+	"#5fd7af",
+	"#5fd7d7",
+	"#5fd7ff",
+	"#5fff00",
+	"#5fff5f",
+	"#5fff87",
+	"#5fffaf",
+	"#5fffd7",
+	"#5fffff",
+	"#870000",
+	"#87005f",
+	"#870087",
+	"#8700af",
+	"#8700d7",
+	"#8700ff",
+	"#875f00",
+	"#875f5f",
+	"#875f87",
+	"#875faf",
+	"#875fd7",
+	"#875fff",
+	"#878700",
+	"#87875f",
+	"#878787",
+	"#8787af",
+	"#8787d7",
+	"#8787ff",
+	"#87af00",
+	"#87af5f",
+	"#87af87",
+	"#87afaf",
+	"#87afd7",
+	"#87afff",
+	"#87d700",
+	"#87d75f",
+	"#87d787",
+	"#87d7af",
+	"#87d7d7",
+	"#87d7ff",
+	"#87ff00",
+	"#87ff5f",
+	"#87ff87",
+	"#87ffaf",
+	"#87ffd7",
+	"#87ffff",
+	"#af0000",
+	"#af005f",
+	"#af0087",
+	"#af00af",
+	"#af00d7",
+	"#af00ff",
+	"#af5f00",
+	"#af5f5f",
+	"#af5f87",
+	"#af5faf",
+	"#af5fd7",
+	"#af5fff",
+	"#af8700",
+	"#af875f",
+	"#af8787",
+	"#af87af",
+	"#af87d7",
+	"#af87ff",
+	"#afaf00",
+	"#afaf5f",
+	"#afaf87",
+	"#afafaf",
+	"#afafd7",
+	"#afafff",
+	"#afd700",
+	"#afd75f",
+	"#afd787",
+	"#afd7af",
+	"#afd7d7",
+	"#afd7ff",
+	"#afff00",
+	"#afff5f",
+	"#afff87",
+	"#afffaf",
+	"#afffd7",
+	"#afffff",
+	"#d70000",
+	"#d7005f",
+	"#d70087",
+	"#d700af",
+	"#d700d7",
+	"#d700ff",
+	"#d75f00",
+	"#d75f5f",
+	"#d75f87",
+	"#d75faf",
+	"#d75fd7",
+	"#d75fff",
+	"#d78700",
+	"#d7875f",
+	"#d78787",
+	"#d787af",
+	"#d787d7",
+	"#d787ff",
+	"#d7af00",
+	"#d7af5f",
+	"#d7af87",
+	"#d7afaf",
+	"#d7afd7",
+	"#d7afff",
+	"#d7d700",
+	"#d7d75f",
+	"#d7d787",
+	"#d7d7af",
+	"#d7d7d7",
+	"#d7d7ff",
+	"#d7ff00",
+	"#d7ff5f",
+	"#d7ff87",
+	"#d7ffaf",
+	"#d7ffd7",
+	"#d7ffff",
+	"#ff0000",
+	"#ff005f",
+	"#ff0087",
+	"#ff00af",
+	"#ff00d7",
+	"#ff00ff",
+	"#ff5f00",
+	"#ff5f5f",
+	"#ff5f87",
+	"#ff5faf",
+	"#ff5fd7",
+	"#ff5fff",
+	"#ff8700",
+	"#ff875f",
+	"#ff8787",
+	"#ff87af",
+	"#ff87d7",
+	"#ff87ff",
+	"#ffaf00",
+	"#ffaf5f",
+	"#ffaf87",
+	"#ffafaf",
+	"#ffafd7",
+	"#ffafff",
+	"#ffd700",
+	"#ffd75f",
+	"#ffd787",
+	"#ffd7af",
+	"#ffd7d7",
+	"#ffd7ff",
+	"#ffff00",
+	"#ffff5f",
+	"#ffff87",
+	"#ffffaf",
+	"#ffffd7",
+	"#ffffff",
+	"#080808",
+	"#121212",
+	"#1c1c1c",
+	"#262626",
+	"#303030",
+	"#3a3a3a",
+	"#444444",
+	"#4e4e4e",
+	"#585858",
+	"#626262",
+	"#6c6c6c",
+	"#767676",
+	"#808080",
+	"#8a8a8a",
+	"#949494",
+	"#9e9e9e",
+	"#a8a8a8",
+	"#b2b2b2",
+	"#bcbcbc",
+	"#c6c6c6",
+	"#d0d0d0",
+	"#dadada",
+	"#e4e4e4",
+	"#eeeeee",
+}
diff --git a/vendor/github.com/charmbracelet/colorprofile/writer.go b/vendor/github.com/charmbracelet/colorprofile/writer.go
new file mode 100644
index 0000000..d04b3b9
--- /dev/null
+++ b/vendor/github.com/charmbracelet/colorprofile/writer.go
@@ -0,0 +1,166 @@
+package colorprofile
+
+import (
+	"bytes"
+	"image/color"
+	"io"
+	"strconv"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// NewWriter creates a new color profile writer that downgrades color sequences
+// based on the detected color profile.
+//
+// If environ is nil, it will use os.Environ() to get the environment variables.
+//
+// It queries the given writer to determine if it supports ANSI escape codes.
+// If it does, along with the given environment variables, it will determine
+// the appropriate color profile to use for color formatting.
+//
+// This respects the NO_COLOR, CLICOLOR, and CLICOLOR_FORCE environment variables.
+func NewWriter(w io.Writer, environ []string) *Writer {
+	return &Writer{
+		Forward: w,
+		Profile: Detect(w, environ),
+	}
+}
+
+// Writer represents a color profile writer that writes ANSI sequences to the
+// underlying writer.
+type Writer struct {
+	Forward io.Writer
+	Profile Profile
+}
+
+// Write writes the given text to the underlying writer.
+func (w *Writer) Write(p []byte) (int, error) {
+	switch w.Profile {
+	case TrueColor:
+		return w.Forward.Write(p)
+	case NoTTY:
+		return io.WriteString(w.Forward, ansi.Strip(string(p)))
+	default:
+		return w.downsample(p)
+	}
+}
+
+// downsample downgrades the given text to the appropriate color profile.
+func (w *Writer) downsample(p []byte) (int, error) {
+	var buf bytes.Buffer
+	var state byte
+
+	parser := ansi.GetParser()
+	defer ansi.PutParser(parser)
+
+	for len(p) > 0 {
+		parser.Reset()
+		seq, _, read, newState := ansi.DecodeSequence(p, state, parser)
+
+		switch {
+		case ansi.HasCsiPrefix(seq) && parser.Command() == 'm':
+			handleSgr(w, parser, &buf)
+		default:
+			// If we're not a style SGR sequence, just write the bytes.
+			if n, err := buf.Write(seq); err != nil {
+				return n, err
+			}
+		}
+
+		p = p[read:]
+		state = newState
+	}
+
+	return w.Forward.Write(buf.Bytes())
+}
+
+// WriteString writes the given text to the underlying writer.
+func (w *Writer) WriteString(s string) (n int, err error) {
+	return w.Write([]byte(s))
+}
+
+func handleSgr(w *Writer, p *ansi.Parser, buf *bytes.Buffer) {
+	var style ansi.Style
+	params := p.Params()
+	for i := 0; i < len(params); i++ {
+		param := params[i]
+
+		switch param := param.Param(0); param {
+		case 0:
+			// SGR default parameter is 0. We use an empty string to reduce the
+			// number of bytes written to the buffer.
+			style = append(style, "")
+		case 30, 31, 32, 33, 34, 35, 36, 37: // 8-bit foreground color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.ForegroundColor(
+				w.Profile.Convert(ansi.BasicColor(param - 30))) //nolint:gosec
+		case 38: // 16 or 24-bit foreground color
+			var c color.Color
+			if n := ansi.ReadStyleColor(params[i:], &c); n > 0 {
+				i += n - 1
+			}
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.ForegroundColor(w.Profile.Convert(c))
+		case 39: // default foreground color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.DefaultForegroundColor()
+		case 40, 41, 42, 43, 44, 45, 46, 47: // 8-bit background color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.BackgroundColor(
+				w.Profile.Convert(ansi.BasicColor(param - 40))) //nolint:gosec
+		case 48: // 16 or 24-bit background color
+			var c color.Color
+			if n := ansi.ReadStyleColor(params[i:], &c); n > 0 {
+				i += n - 1
+			}
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.BackgroundColor(w.Profile.Convert(c))
+		case 49: // default background color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.DefaultBackgroundColor()
+		case 58: // 16 or 24-bit underline color
+			var c color.Color
+			if n := ansi.ReadStyleColor(params[i:], &c); n > 0 {
+				i += n - 1
+			}
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.UnderlineColor(w.Profile.Convert(c))
+		case 59: // default underline color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.DefaultUnderlineColor()
+		case 90, 91, 92, 93, 94, 95, 96, 97: // 8-bit bright foreground color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.ForegroundColor(
+				w.Profile.Convert(ansi.BasicColor(param - 90 + 8))) //nolint:gosec
+		case 100, 101, 102, 103, 104, 105, 106, 107: // 8-bit bright background color
+			if w.Profile < ANSI {
+				continue
+			}
+			style = style.BackgroundColor(
+				w.Profile.Convert(ansi.BasicColor(param - 100 + 8))) //nolint:gosec
+		default:
+			// If this is not a color attribute, just append it to the style.
+			style = append(style, strconv.Itoa(param))
+		}
+	}
+
+	_, _ = buf.WriteString(style.String())
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/.gitignore b/vendor/github.com/charmbracelet/lipgloss/.gitignore
new file mode 100644
index 0000000..db48201
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/.gitignore
@@ -0,0 +1,2 @@
+ssh_example_ed25519*
+dist/
diff --git a/vendor/github.com/charmbracelet/lipgloss/.golangci.yml b/vendor/github.com/charmbracelet/lipgloss/.golangci.yml
new file mode 100644
index 0000000..90c5c08
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/.golangci.yml
@@ -0,0 +1,41 @@
+run:
+  tests: false
+
+issues:
+  include:
+    - EXC0001
+    - EXC0005
+    - EXC0011
+    - EXC0012
+    - EXC0013
+
+  max-issues-per-linter: 0
+  max-same-issues: 0
+
+linters:
+  enable:
+    - bodyclose
+    - exhaustive
+    - goconst
+    - godot
+    - godox
+    - gofumpt
+    - goimports
+    - gomoddirectives
+    - goprintffuncname
+    - gosec
+    - misspell
+    - nakedret
+    - nestif
+    - nilerr
+    - noctx
+    - nolintlint
+    - prealloc
+    - revive
+    - rowserrcheck
+    - sqlclosecheck
+    - tparallel
+    - unconvert
+    - unparam
+    - whitespace
+    - wrapcheck
diff --git a/vendor/github.com/charmbracelet/lipgloss/.goreleaser.yml b/vendor/github.com/charmbracelet/lipgloss/.goreleaser.yml
new file mode 100644
index 0000000..3353d02
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/.goreleaser.yml
@@ -0,0 +1,5 @@
+# yaml-language-server: $schema=https://goreleaser.com/static/schema-pro.json
+version: 2
+includes:
+  - from_url:
+      url: charmbracelet/meta/main/goreleaser-lib.yaml
diff --git a/vendor/github.com/charmbracelet/lipgloss/LICENSE b/vendor/github.com/charmbracelet/lipgloss/LICENSE
new file mode 100644
index 0000000..6f5b1fa
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2021-2023 Charmbracelet, Inc
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/vendor/github.com/charmbracelet/lipgloss/README.md b/vendor/github.com/charmbracelet/lipgloss/README.md
new file mode 100644
index 0000000..cee2371
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/README.md
@@ -0,0 +1,815 @@
+# Lip Gloss
+
+<p>
+    <a href="https://stuff.charm.sh/lipgloss/lipgloss-mascot-2k.png"><img width="340" alt="Lip Gloss title treatment" src="https://github.com/charmbracelet/lipgloss/assets/25087/147cadb1-4254-43ec-ae6b-8d6ca7b029a1"></a><br>
+    <a href="https://github.com/charmbracelet/lipgloss/releases"><img src="https://img.shields.io/github/release/charmbracelet/lipgloss.svg" alt="Latest Release"></a>
+    <a href="https://pkg.go.dev/github.com/charmbracelet/lipgloss?tab=doc"><img src="https://godoc.org/github.com/golang/gddo?status.svg" alt="GoDoc"></a>
+    <a href="https://github.com/charmbracelet/lipgloss/actions"><img src="https://github.com/charmbracelet/lipgloss/workflows/build/badge.svg" alt="Build Status"></a>
+    <a href="https://www.phorm.ai/query?projectId=a0e324b6-b706-4546-b951-6671ea60c13f"><img src="https://stuff.charm.sh/misc/phorm-badge.svg" alt="phorm.ai"></a>
+</p>
+
+Style definitions for nice terminal layouts. Built with TUIs in mind.
+
+![Lip Gloss example](https://github.com/user-attachments/assets/7950b1c1-e0e3-427e-8e7d-6f7f6ad17ca7)
+
+Lip Gloss takes an expressive, declarative approach to terminal rendering.
+Users familiar with CSS will feel at home with Lip Gloss.
+
+```go
+
+import "github.com/charmbracelet/lipgloss"
+
+var style = lipgloss.NewStyle().
+    Bold(true).
+    Foreground(lipgloss.Color("#FAFAFA")).
+    Background(lipgloss.Color("#7D56F4")).
+    PaddingTop(2).
+    PaddingLeft(4).
+    Width(22)
+
+fmt.Println(style.Render("Hello, kitty"))
+```
+
+## Colors
+
+Lip Gloss supports the following color profiles:
+
+### ANSI 16 colors (4-bit)
+
+```go
+lipgloss.Color("5")  // magenta
+lipgloss.Color("9")  // red
+lipgloss.Color("12") // light blue
+```
+
+### ANSI 256 Colors (8-bit)
+
+```go
+lipgloss.Color("86")  // aqua
+lipgloss.Color("201") // hot pink
+lipgloss.Color("202") // orange
+```
+
+### True Color (16,777,216 colors; 24-bit)
+
+```go
+lipgloss.Color("#0000FF") // good ol' 100% blue
+lipgloss.Color("#04B575") // a green
+lipgloss.Color("#3C3C3C") // a dark gray
+```
+
+...as well as a 1-bit ASCII profile, which is black and white only.
+
+The terminal's color profile will be automatically detected, and colors outside
+the gamut of the current palette will be automatically coerced to their closest
+available value.
+
+### Adaptive Colors
+
+You can also specify color options for light and dark backgrounds:
+
+```go
+lipgloss.AdaptiveColor{Light: "236", Dark: "248"}
+```
+
+The terminal's background color will automatically be detected and the
+appropriate color will be chosen at runtime.
+
+### Complete Colors
+
+CompleteColor specifies exact values for True Color, ANSI256, and ANSI color
+profiles.
+
+```go
+lipgloss.CompleteColor{TrueColor: "#0000FF", ANSI256: "86", ANSI: "5"}
+```
+
+Automatic color degradation will not be performed in this case and it will be
+based on the color specified.
+
+### Complete Adaptive Colors
+
+You can use `CompleteColor` with `AdaptiveColor` to specify the exact values for
+light and dark backgrounds without automatic color degradation.
+
+```go
+lipgloss.CompleteAdaptiveColor{
+    Light: CompleteColor{TrueColor: "#d7ffae", ANSI256: "193", ANSI: "11"},
+    Dark:  CompleteColor{TrueColor: "#d75fee", ANSI256: "163", ANSI: "5"},
+}
+```
+
+## Inline Formatting
+
+Lip Gloss supports the usual ANSI text formatting options:
+
+```go
+var style = lipgloss.NewStyle().
+    Bold(true).
+    Italic(true).
+    Faint(true).
+    Blink(true).
+    Strikethrough(true).
+    Underline(true).
+    Reverse(true)
+```
+
+## Block-Level Formatting
+
+Lip Gloss also supports rules for block-level formatting:
+
+```go
+// Padding
+var style = lipgloss.NewStyle().
+    PaddingTop(2).
+    PaddingRight(4).
+    PaddingBottom(2).
+    PaddingLeft(4)
+
+// Margins
+var style = lipgloss.NewStyle().
+    MarginTop(2).
+    MarginRight(4).
+    MarginBottom(2).
+    MarginLeft(4)
+```
+
+There is also shorthand syntax for margins and padding, which follows the same
+format as CSS:
+
+```go
+// 2 cells on all sides
+lipgloss.NewStyle().Padding(2)
+
+// 2 cells on the top and bottom, 4 cells on the left and right
+lipgloss.NewStyle().Margin(2, 4)
+
+// 1 cell on the top, 4 cells on the sides, 2 cells on the bottom
+lipgloss.NewStyle().Padding(1, 4, 2)
+
+// Clockwise, starting from the top: 2 cells on the top, 4 on the right, 3 on
+// the bottom, and 1 on the left
+lipgloss.NewStyle().Margin(2, 4, 3, 1)
+```
+
+## Aligning Text
+
+You can align paragraphs of text to the left, right, or center.
+
+```go
+var style = lipgloss.NewStyle().
+    Width(24).
+    Align(lipgloss.Left).  // align it left
+    Align(lipgloss.Right). // no wait, align it right
+    Align(lipgloss.Center) // just kidding, align it in the center
+```
+
+## Width and Height
+
+Setting a minimum width and height is simple and straightforward.
+
+```go
+var style = lipgloss.NewStyle().
+    SetString("What’s for lunch?").
+    Width(24).
+    Height(32).
+    Foreground(lipgloss.Color("63"))
+```
+
+## Borders
+
+Adding borders is easy:
+
+```go
+// Add a purple, rectangular border
+var style = lipgloss.NewStyle().
+    BorderStyle(lipgloss.NormalBorder()).
+    BorderForeground(lipgloss.Color("63"))
+
+// Set a rounded, yellow-on-purple border to the top and left
+var anotherStyle = lipgloss.NewStyle().
+    BorderStyle(lipgloss.RoundedBorder()).
+    BorderForeground(lipgloss.Color("228")).
+    BorderBackground(lipgloss.Color("63")).
+    BorderTop(true).
+    BorderLeft(true)
+
+// Make your own border
+var myCuteBorder = lipgloss.Border{
+    Top:         "._.:*:",
+    Bottom:      "._.:*:",
+    Left:        "|*",
+    Right:       "|*",
+    TopLeft:     "*",
+    TopRight:    "*",
+    BottomLeft:  "*",
+    BottomRight: "*",
+}
+```
+
+There are also shorthand functions for defining borders, which follow a similar
+pattern to the margin and padding shorthand functions.
+
+```go
+// Add a thick border to the top and bottom
+lipgloss.NewStyle().
+    Border(lipgloss.ThickBorder(), true, false)
+
+// Add a double border to the top and left sides. Rules are set clockwise
+// from top.
+lipgloss.NewStyle().
+    Border(lipgloss.DoubleBorder(), true, false, false, true)
+```
+
+For more on borders see [the docs][docs].
+
+## Copying Styles
+
+Just use assignment:
+
+```go
+style := lipgloss.NewStyle().Foreground(lipgloss.Color("219"))
+
+copiedStyle := style // this is a true copy
+
+wildStyle := style.Blink(true) // this is also true copy, with blink added
+
+```
+
+Since `Style` data structures contains only primitive types, assigning a style
+to another effectively creates a new copy of the style without mutating the
+original.
+
+## Inheritance
+
+Styles can inherit rules from other styles. When inheriting, only unset rules
+on the receiver are inherited.
+
+```go
+var styleA = lipgloss.NewStyle().
+    Foreground(lipgloss.Color("229")).
+    Background(lipgloss.Color("63"))
+
+// Only the background color will be inherited here, because the foreground
+// color will have been already set:
+var styleB = lipgloss.NewStyle().
+    Foreground(lipgloss.Color("201")).
+    Inherit(styleA)
+```
+
+## Unsetting Rules
+
+All rules can be unset:
+
+```go
+var style = lipgloss.NewStyle().
+    Bold(true).                        // make it bold
+    UnsetBold().                       // jk don't make it bold
+    Background(lipgloss.Color("227")). // yellow background
+    UnsetBackground()                  // never mind
+```
+
+When a rule is unset, it won't be inherited or copied.
+
+## Enforcing Rules
+
+Sometimes, such as when developing a component, you want to make sure style
+definitions respect their intended purpose in the UI. This is where `Inline`
+and `MaxWidth`, and `MaxHeight` come in:
+
+```go
+// Force rendering onto a single line, ignoring margins, padding, and borders.
+someStyle.Inline(true).Render("yadda yadda")
+
+// Also limit rendering to five cells
+someStyle.Inline(true).MaxWidth(5).Render("yadda yadda")
+
+// Limit rendering to a 5x5 cell block
+someStyle.MaxWidth(5).MaxHeight(5).Render("yadda yadda")
+```
+
+## Tabs
+
+The tab character (`\t`) is rendered differently in different terminals (often
+as 8 spaces, sometimes 4). Because of this inconsistency, Lip Gloss converts
+tabs to 4 spaces at render time. This behavior can be changed on a per-style
+basis, however:
+
+```go
+style := lipgloss.NewStyle() // tabs will render as 4 spaces, the default
+style = style.TabWidth(2)    // render tabs as 2 spaces
+style = style.TabWidth(0)    // remove tabs entirely
+style = style.TabWidth(lipgloss.NoTabConversion) // leave tabs intact
+```
+
+## Rendering
+
+Generally, you just call the `Render(string...)` method on a `lipgloss.Style`:
+
+```go
+style := lipgloss.NewStyle().Bold(true).SetString("Hello,")
+fmt.Println(style.Render("kitty.")) // Hello, kitty.
+fmt.Println(style.Render("puppy.")) // Hello, puppy.
+```
+
+But you could also use the Stringer interface:
+
+```go
+var style = lipgloss.NewStyle().SetString("你好，猫咪。").Bold(true)
+fmt.Println(style) // 你好，猫咪。
+```
+
+### Custom Renderers
+
+Custom renderers allow you to render to a specific outputs. This is
+particularly important when you want to render to different outputs and
+correctly detect the color profile and dark background status for each, such as
+in a server-client situation.
+
+```go
+func myLittleHandler(sess ssh.Session) {
+    // Create a renderer for the client.
+    renderer := lipgloss.NewRenderer(sess)
+
+    // Create a new style on the renderer.
+    style := renderer.NewStyle().Background(lipgloss.AdaptiveColor{Light: "63", Dark: "228"})
+
+    // Render. The color profile and dark background state will be correctly detected.
+    io.WriteString(sess, style.Render("Heyyyyyyy"))
+}
+```
+
+For an example on using a custom renderer over SSH with [Wish][wish] see the
+[SSH example][ssh-example].
+
+## Utilities
+
+In addition to pure styling, Lip Gloss also ships with some utilities to help
+assemble your layouts.
+
+### Joining Paragraphs
+
+Horizontally and vertically joining paragraphs is a cinch.
+
+```go
+// Horizontally join three paragraphs along their bottom edges
+lipgloss.JoinHorizontal(lipgloss.Bottom, paragraphA, paragraphB, paragraphC)
+
+// Vertically join two paragraphs along their center axes
+lipgloss.JoinVertical(lipgloss.Center, paragraphA, paragraphB)
+
+// Horizontally join three paragraphs, with the shorter ones aligning 20%
+// from the top of the tallest
+lipgloss.JoinHorizontal(0.2, paragraphA, paragraphB, paragraphC)
+```
+
+### Measuring Width and Height
+
+Sometimes you’ll want to know the width and height of text blocks when building
+your layouts.
+
+```go
+// Render a block of text.
+var style = lipgloss.NewStyle().
+    Width(40).
+    Padding(2)
+var block string = style.Render(someLongString)
+
+// Get the actual, physical dimensions of the text block.
+width := lipgloss.Width(block)
+height := lipgloss.Height(block)
+
+// Here's a shorthand function.
+w, h := lipgloss.Size(block)
+```
+
+### Placing Text in Whitespace
+
+Sometimes you’ll simply want to place a block of text in whitespace.
+
+```go
+// Center a paragraph horizontally in a space 80 cells wide. The height of
+// the block returned will be as tall as the input paragraph.
+block := lipgloss.PlaceHorizontal(80, lipgloss.Center, fancyStyledParagraph)
+
+// Place a paragraph at the bottom of a space 30 cells tall. The width of
+// the text block returned will be as wide as the input paragraph.
+block := lipgloss.PlaceVertical(30, lipgloss.Bottom, fancyStyledParagraph)
+
+// Place a paragraph in the bottom right corner of a 30x80 cell space.
+block := lipgloss.Place(30, 80, lipgloss.Right, lipgloss.Bottom, fancyStyledParagraph)
+```
+
+You can also style the whitespace. For details, see [the docs][docs].
+
+## Rendering Tables
+
+Lip Gloss ships with a table rendering sub-package.
+
+```go
+import "github.com/charmbracelet/lipgloss/table"
+```
+
+Define some rows of data.
+
+```go
+rows := [][]string{
+    {"Chinese", "您好", "你好"},
+    {"Japanese", "こんにちは", "やあ"},
+    {"Arabic", "أهلين", "أهلا"},
+    {"Russian", "Здравствуйте", "Привет"},
+    {"Spanish", "Hola", "¿Qué tal?"},
+}
+```
+
+Use the table package to style and render the table.
+
+```go
+var (
+    purple    = lipgloss.Color("99")
+    gray      = lipgloss.Color("245")
+    lightGray = lipgloss.Color("241")
+
+    headerStyle  = lipgloss.NewStyle().Foreground(purple).Bold(true).Align(lipgloss.Center)
+    cellStyle    = lipgloss.NewStyle().Padding(0, 1).Width(14)
+    oddRowStyle  = cellStyle.Foreground(gray)
+    evenRowStyle = cellStyle.Foreground(lightGray)
+)
+
+t := table.New().
+    Border(lipgloss.NormalBorder()).
+    BorderStyle(lipgloss.NewStyle().Foreground(purple)).
+    StyleFunc(func(row, col int) lipgloss.Style {
+        switch {
+        case row == table.HeaderRow:
+            return headerStyle
+        case row%2 == 0:
+            return evenRowStyle
+        default:
+            return oddRowStyle
+        }
+    }).
+    Headers("LANGUAGE", "FORMAL", "INFORMAL").
+    Rows(rows...)
+
+// You can also add tables row-by-row
+t.Row("English", "You look absolutely fabulous.", "How's it going?")
+```
+
+Print the table.
+
+```go
+fmt.Println(t)
+```
+
+![Table Example](https://github.com/charmbracelet/lipgloss/assets/42545625/6e4b70c4-f494-45da-a467-bdd27df30d5d)
+
+> [!WARNING]
+> Table `Rows` need to be declared before `Offset` otherwise it does nothing.
+
+### Table Borders
+
+There are helpers to generate tables in markdown or ASCII style:
+
+#### Markdown Table
+
+```go
+table.New().Border(lipgloss.MarkdownBorder()).BorderTop(false).BorderBottom(false)
+```
+
+```
+| LANGUAGE |    FORMAL    | INFORMAL  |
+|----------|--------------|-----------|
+| Chinese  | Nǐn hǎo      | Nǐ hǎo    |
+| French   | Bonjour      | Salut     |
+| Russian  | Zdravstvuyte | Privet    |
+| Spanish  | Hola         | ¿Qué tal? |
+```
+
+#### ASCII Table
+
+```go
+table.New().Border(lipgloss.ASCIIBorder())
+```
+
+```
++----------+--------------+-----------+
+| LANGUAGE |    FORMAL    | INFORMAL  |
++----------+--------------+-----------+
+| Chinese  | Nǐn hǎo      | Nǐ hǎo    |
+| French   | Bonjour      | Salut     |
+| Russian  | Zdravstvuyte | Privet    |
+| Spanish  | Hola         | ¿Qué tal? |
++----------+--------------+-----------+
+```
+
+For more on tables see [the docs](https://pkg.go.dev/github.com/charmbracelet/lipgloss?tab=doc) and [examples](https://github.com/charmbracelet/lipgloss/tree/master/examples/table).
+
+## Rendering Lists
+
+Lip Gloss ships with a list rendering sub-package.
+
+```go
+import "github.com/charmbracelet/lipgloss/list"
+```
+
+Define a new list.
+
+```go
+l := list.New("A", "B", "C")
+```
+
+Print the list.
+
+```go
+fmt.Println(l)
+
+// • A
+// • B
+// • C
+```
+
+Lists have the ability to nest.
+
+```go
+l := list.New(
+    "A", list.New("Artichoke"),
+    "B", list.New("Baking Flour", "Bananas", "Barley", "Bean Sprouts"),
+    "C", list.New("Cashew Apple", "Cashews", "Coconut Milk", "Curry Paste", "Currywurst"),
+    "D", list.New("Dill", "Dragonfruit", "Dried Shrimp"),
+    "E", list.New("Eggs"),
+    "F", list.New("Fish Cake", "Furikake"),
+    "J", list.New("Jicama"),
+    "K", list.New("Kohlrabi"),
+    "L", list.New("Leeks", "Lentils", "Licorice Root"),
+)
+```
+
+Print the list.
+
+```go
+fmt.Println(l)
+```
+
+<p align="center">
+<img width="600" alt="image" src="https://github.com/charmbracelet/lipgloss/assets/42545625/0dc9f440-0748-4151-a3b0-7dcf29dfcdb0">
+</p>
+
+Lists can be customized via their enumeration function as well as using
+`lipgloss.Style`s.
+
+```go
+enumeratorStyle := lipgloss.NewStyle().Foreground(lipgloss.Color("99")).MarginRight(1)
+itemStyle := lipgloss.NewStyle().Foreground(lipgloss.Color("212")).MarginRight(1)
+
+l := list.New(
+    "Glossier",
+    "Claire’s Boutique",
+    "Nyx",
+    "Mac",
+    "Milk",
+    ).
+    Enumerator(list.Roman).
+    EnumeratorStyle(enumeratorStyle).
+    ItemStyle(itemStyle)
+```
+
+Print the list.
+
+<p align="center">
+<img width="600" alt="List example" src="https://github.com/charmbracelet/lipgloss/assets/42545625/360494f1-57fb-4e13-bc19-0006efe01561">
+</p>
+
+In addition to the predefined enumerators (`Arabic`, `Alphabet`, `Roman`, `Bullet`, `Tree`),
+you may also define your own custom enumerator:
+
+```go
+l := list.New("Duck", "Duck", "Duck", "Duck", "Goose", "Duck", "Duck")
+
+func DuckDuckGooseEnumerator(l list.Items, i int) string {
+    if l.At(i).Value() == "Goose" {
+        return "Honk →"
+    }
+    return ""
+}
+
+l = l.Enumerator(DuckDuckGooseEnumerator)
+```
+
+Print the list:
+
+<p align="center">
+<img width="600" alt="image" src="https://github.com/charmbracelet/lipgloss/assets/42545625/157aaf30-140d-4948-9bb4-dfba46e5b87e">
+</p>
+
+If you need, you can also build lists incrementally:
+
+```go
+l := list.New()
+
+for i := 0; i < repeat; i++ {
+    l.Item("Lip Gloss")
+}
+```
+
+## Rendering Trees
+
+Lip Gloss ships with a tree rendering sub-package.
+
+```go
+import "github.com/charmbracelet/lipgloss/tree"
+```
+
+Define a new tree.
+
+```go
+t := tree.Root(".").
+    Child("A", "B", "C")
+```
+
+Print the tree.
+
+```go
+fmt.Println(t)
+
+// .
+// ├── A
+// ├── B
+// └── C
+```
+
+Trees have the ability to nest.
+
+```go
+t := tree.Root(".").
+    Child("macOS").
+    Child(
+        tree.New().
+            Root("Linux").
+            Child("NixOS").
+            Child("Arch Linux (btw)").
+            Child("Void Linux"),
+        ).
+    Child(
+        tree.New().
+            Root("BSD").
+            Child("FreeBSD").
+            Child("OpenBSD"),
+    )
+```
+
+Print the tree.
+
+```go
+fmt.Println(t)
+```
+
+<p align="center">
+<img width="663" alt="Tree Example (simple)" src="https://github.com/user-attachments/assets/5ef14eb8-a5d4-4f94-8834-e15d1e714f89">
+</p>
+
+Trees can be customized via their enumeration function as well as using
+`lipgloss.Style`s.
+
+```go
+enumeratorStyle := lipgloss.NewStyle().Foreground(lipgloss.Color("63")).MarginRight(1)
+rootStyle := lipgloss.NewStyle().Foreground(lipgloss.Color("35"))
+itemStyle := lipgloss.NewStyle().Foreground(lipgloss.Color("212"))
+
+t := tree.
+    Root("⁜ Makeup").
+    Child(
+        "Glossier",
+        "Fenty Beauty",
+        tree.New().Child(
+            "Gloss Bomb Universal Lip Luminizer",
+            "Hot Cheeks Velour Blushlighter",
+        ),
+        "Nyx",
+        "Mac",
+        "Milk",
+    ).
+    Enumerator(tree.RoundedEnumerator).
+    EnumeratorStyle(enumeratorStyle).
+    RootStyle(rootStyle).
+    ItemStyle(itemStyle)
+```
+
+Print the tree.
+
+<p align="center">
+<img width="663" alt="Tree Example (makeup)" src="https://github.com/user-attachments/assets/06d12d87-744a-4c89-bd98-45de9094a97e">
+</p>
+
+The predefined enumerators for trees are `DefaultEnumerator` and `RoundedEnumerator`.
+
+If you need, you can also build trees incrementally:
+
+```go
+t := tree.New()
+
+for i := 0; i < repeat; i++ {
+    t.Child("Lip Gloss")
+}
+```
+
+---
+
+## FAQ
+
+<details>
+<summary>
+Why are things misaligning? Why are borders at the wrong widths?
+</summary>
+<p>This is most likely due to your locale and encoding, particularly with
+regard to Chinese, Japanese, and Korean (for example, <code>zh_CN.UTF-8</code>
+or <code>ja_JP.UTF-8</code>). The most direct way to fix this is to set
+<code>RUNEWIDTH_EASTASIAN=0</code> in your environment.</p>
+
+<p>For details see <a href="https://github.com/charmbracelet/lipgloss/issues/40">https://github.com/charmbracelet/lipgloss/issues/40.</a></p>
+</details>
+
+<details>
+<summary>
+Why isn't Lip Gloss displaying colors?
+</summary>
+<p>Lip Gloss automatically degrades colors to the best available option in the
+given terminal, and if output's not a TTY it will remove color output entirely.
+This is common when running tests, CI, or when piping output elsewhere.</p>
+
+<p>If necessary, you can force a color profile in your tests with
+<a href="https://pkg.go.dev/github.com/charmbracelet/lipgloss#SetColorProfile"><code>SetColorProfile</code></a>.</p>
+
+```go
+import (
+    "github.com/charmbracelet/lipgloss"
+    "github.com/muesli/termenv"
+)
+
+lipgloss.SetColorProfile(termenv.TrueColor)
+```
+
+_Note:_ this option limits the flexibility of your application and can cause
+ANSI escape codes to be output in cases where that might not be desired. Take
+careful note of your use case and environment before choosing to force a color
+profile.
+
+</details>
+
+## What about [Bubble Tea][tea]?
+
+Lip Gloss doesn’t replace Bubble Tea. Rather, it is an excellent Bubble Tea
+companion. It was designed to make assembling terminal user interface views as
+simple and fun as possible so that you can focus on building your application
+instead of concerning yourself with low-level layout details.
+
+In simple terms, you can use Lip Gloss to help build your Bubble Tea views.
+
+[tea]: https://github.com/charmbracelet/tea
+
+## Under the Hood
+
+Lip Gloss is built on the excellent [Termenv][termenv] and [Reflow][reflow]
+libraries which deal with color and ANSI-aware text operations, respectively.
+For many use cases Termenv and Reflow will be sufficient for your needs.
+
+[termenv]: https://github.com/muesli/termenv
+[reflow]: https://github.com/muesli/reflow
+
+## Rendering Markdown
+
+For a more document-centric rendering solution with support for things like
+lists, tables, and syntax-highlighted code have a look at [Glamour][glamour],
+the stylesheet-based Markdown renderer.
+
+[glamour]: https://github.com/charmbracelet/glamour
+
+## Contributing
+
+See [contributing][contribute].
+
+[contribute]: https://github.com/charmbracelet/lipgloss/contribute
+
+## Feedback
+
+We’d love to hear your thoughts on this project. Feel free to drop us a note!
+
+- [Twitter](https://twitter.com/charmcli)
+- [The Fediverse](https://mastodon.social/@charmcli)
+- [Discord](https://charm.sh/chat)
+
+## License
+
+[MIT](https://github.com/charmbracelet/lipgloss/raw/master/LICENSE)
+
+---
+
+Part of [Charm](https://charm.sh).
+
+<a href="https://charm.sh/"><img alt="The Charm logo" src="https://stuff.charm.sh/charm-badge.jpg" width="400"></a>
+
+Charm热爱开源 • Charm loves open source
+
+[docs]: https://pkg.go.dev/github.com/charmbracelet/lipgloss?tab=doc
+[wish]: https://github.com/charmbracelet/wish
+[ssh-example]: examples/ssh
diff --git a/vendor/github.com/charmbracelet/lipgloss/Taskfile.yaml b/vendor/github.com/charmbracelet/lipgloss/Taskfile.yaml
new file mode 100644
index 0000000..0b4a771
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/Taskfile.yaml
@@ -0,0 +1,19 @@
+# https://taskfile.dev
+
+version: '3'
+
+tasks:
+  lint:
+    desc: Run base linters
+    cmds:
+      - golangci-lint run
+
+  test:
+    desc: Run tests
+    cmds:
+      - go test ./... {{.CLI_ARGS}}
+
+  test:table:
+    desc: Run table tests
+    cmds:
+      - go test ./table {{.CLI_ARGS}}
diff --git a/vendor/github.com/charmbracelet/lipgloss/align.go b/vendor/github.com/charmbracelet/lipgloss/align.go
new file mode 100644
index 0000000..ce654b2
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/align.go
@@ -0,0 +1,83 @@
+package lipgloss
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+	"github.com/muesli/termenv"
+)
+
+// Perform text alignment. If the string is multi-lined, we also make all lines
+// the same width by padding them with spaces. If a termenv style is passed,
+// use that to style the spaces added.
+func alignTextHorizontal(str string, pos Position, width int, style *termenv.Style) string {
+	lines, widestLine := getLines(str)
+	var b strings.Builder
+
+	for i, l := range lines {
+		lineWidth := ansi.StringWidth(l)
+
+		shortAmount := widestLine - lineWidth                // difference from the widest line
+		shortAmount += max(0, width-(shortAmount+lineWidth)) // difference from the total width, if set
+
+		if shortAmount > 0 {
+			switch pos { //nolint:exhaustive
+			case Right:
+				s := strings.Repeat(" ", shortAmount)
+				if style != nil {
+					s = style.Styled(s)
+				}
+				l = s + l
+			case Center:
+				// Note: remainder goes on the right.
+				left := shortAmount / 2       //nolint:mnd
+				right := left + shortAmount%2 //nolint:mnd
+
+				leftSpaces := strings.Repeat(" ", left)
+				rightSpaces := strings.Repeat(" ", right)
+
+				if style != nil {
+					leftSpaces = style.Styled(leftSpaces)
+					rightSpaces = style.Styled(rightSpaces)
+				}
+				l = leftSpaces + l + rightSpaces
+			default: // Left
+				s := strings.Repeat(" ", shortAmount)
+				if style != nil {
+					s = style.Styled(s)
+				}
+				l += s
+			}
+		}
+
+		b.WriteString(l)
+		if i < len(lines)-1 {
+			b.WriteRune('\n')
+		}
+	}
+
+	return b.String()
+}
+
+func alignTextVertical(str string, pos Position, height int, _ *termenv.Style) string {
+	strHeight := strings.Count(str, "\n") + 1
+	if height < strHeight {
+		return str
+	}
+
+	switch pos {
+	case Top:
+		return str + strings.Repeat("\n", height-strHeight)
+	case Center:
+		topPadding, bottomPadding := (height-strHeight)/2, (height-strHeight)/2 //nolint:mnd
+		if strHeight+topPadding+bottomPadding > height {
+			topPadding--
+		} else if strHeight+topPadding+bottomPadding < height {
+			bottomPadding++
+		}
+		return strings.Repeat("\n", topPadding) + str + strings.Repeat("\n", bottomPadding)
+	case Bottom:
+		return strings.Repeat("\n", height-strHeight) + str
+	}
+	return str
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/ansi_unix.go b/vendor/github.com/charmbracelet/lipgloss/ansi_unix.go
new file mode 100644
index 0000000..d416b8c
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/ansi_unix.go
@@ -0,0 +1,7 @@
+//go:build !windows
+// +build !windows
+
+package lipgloss
+
+// enableLegacyWindowsANSI is only needed on Windows.
+func enableLegacyWindowsANSI() {}
diff --git a/vendor/github.com/charmbracelet/lipgloss/ansi_windows.go b/vendor/github.com/charmbracelet/lipgloss/ansi_windows.go
new file mode 100644
index 0000000..0cf56e4
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/ansi_windows.go
@@ -0,0 +1,22 @@
+//go:build windows
+// +build windows
+
+package lipgloss
+
+import (
+	"sync"
+
+	"github.com/muesli/termenv"
+)
+
+var enableANSI sync.Once
+
+// enableANSIColors enables support for ANSI color sequences in the Windows
+// default console (cmd.exe and the PowerShell application). Note that this
+// only works with Windows 10. Also note that Windows Terminal supports colors
+// by default.
+func enableLegacyWindowsANSI() {
+	enableANSI.Do(func() {
+		_, _ = termenv.EnableWindowsANSIConsole()
+	})
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/borders.go b/vendor/github.com/charmbracelet/lipgloss/borders.go
new file mode 100644
index 0000000..b36f874
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/borders.go
@@ -0,0 +1,490 @@
+package lipgloss
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+	"github.com/muesli/termenv"
+	"github.com/rivo/uniseg"
+)
+
+// Border contains a series of values which comprise the various parts of a
+// border.
+type Border struct {
+	Top          string
+	Bottom       string
+	Left         string
+	Right        string
+	TopLeft      string
+	TopRight     string
+	BottomLeft   string
+	BottomRight  string
+	MiddleLeft   string
+	MiddleRight  string
+	Middle       string
+	MiddleTop    string
+	MiddleBottom string
+}
+
+// GetTopSize returns the width of the top border. If borders contain runes of
+// varying widths, the widest rune is returned. If no border exists on the top
+// edge, 0 is returned.
+func (b Border) GetTopSize() int {
+	return getBorderEdgeWidth(b.TopLeft, b.Top, b.TopRight)
+}
+
+// GetRightSize returns the width of the right border. If borders contain
+// runes of varying widths, the widest rune is returned. If no border exists on
+// the right edge, 0 is returned.
+func (b Border) GetRightSize() int {
+	return getBorderEdgeWidth(b.TopRight, b.Right, b.BottomRight)
+}
+
+// GetBottomSize returns the width of the bottom border. If borders contain
+// runes of varying widths, the widest rune is returned. If no border exists on
+// the bottom edge, 0 is returned.
+func (b Border) GetBottomSize() int {
+	return getBorderEdgeWidth(b.BottomLeft, b.Bottom, b.BottomRight)
+}
+
+// GetLeftSize returns the width of the left border. If borders contain runes
+// of varying widths, the widest rune is returned. If no border exists on the
+// left edge, 0 is returned.
+func (b Border) GetLeftSize() int {
+	return getBorderEdgeWidth(b.TopLeft, b.Left, b.BottomLeft)
+}
+
+func getBorderEdgeWidth(borderParts ...string) (maxWidth int) {
+	for _, piece := range borderParts {
+		w := maxRuneWidth(piece)
+		if w > maxWidth {
+			maxWidth = w
+		}
+	}
+	return maxWidth
+}
+
+var (
+	noBorder = Border{}
+
+	normalBorder = Border{
+		Top:          "─",
+		Bottom:       "─",
+		Left:         "│",
+		Right:        "│",
+		TopLeft:      "┌",
+		TopRight:     "┐",
+		BottomLeft:   "└",
+		BottomRight:  "┘",
+		MiddleLeft:   "├",
+		MiddleRight:  "┤",
+		Middle:       "┼",
+		MiddleTop:    "┬",
+		MiddleBottom: "┴",
+	}
+
+	roundedBorder = Border{
+		Top:          "─",
+		Bottom:       "─",
+		Left:         "│",
+		Right:        "│",
+		TopLeft:      "╭",
+		TopRight:     "╮",
+		BottomLeft:   "╰",
+		BottomRight:  "╯",
+		MiddleLeft:   "├",
+		MiddleRight:  "┤",
+		Middle:       "┼",
+		MiddleTop:    "┬",
+		MiddleBottom: "┴",
+	}
+
+	blockBorder = Border{
+		Top:          "█",
+		Bottom:       "█",
+		Left:         "█",
+		Right:        "█",
+		TopLeft:      "█",
+		TopRight:     "█",
+		BottomLeft:   "█",
+		BottomRight:  "█",
+		MiddleLeft:   "█",
+		MiddleRight:  "█",
+		Middle:       "█",
+		MiddleTop:    "█",
+		MiddleBottom: "█",
+	}
+
+	outerHalfBlockBorder = Border{
+		Top:         "▀",
+		Bottom:      "▄",
+		Left:        "▌",
+		Right:       "▐",
+		TopLeft:     "▛",
+		TopRight:    "▜",
+		BottomLeft:  "▙",
+		BottomRight: "▟",
+	}
+
+	innerHalfBlockBorder = Border{
+		Top:         "▄",
+		Bottom:      "▀",
+		Left:        "▐",
+		Right:       "▌",
+		TopLeft:     "▗",
+		TopRight:    "▖",
+		BottomLeft:  "▝",
+		BottomRight: "▘",
+	}
+
+	thickBorder = Border{
+		Top:          "━",
+		Bottom:       "━",
+		Left:         "┃",
+		Right:        "┃",
+		TopLeft:      "┏",
+		TopRight:     "┓",
+		BottomLeft:   "┗",
+		BottomRight:  "┛",
+		MiddleLeft:   "┣",
+		MiddleRight:  "┫",
+		Middle:       "╋",
+		MiddleTop:    "┳",
+		MiddleBottom: "┻",
+	}
+
+	doubleBorder = Border{
+		Top:          "═",
+		Bottom:       "═",
+		Left:         "║",
+		Right:        "║",
+		TopLeft:      "╔",
+		TopRight:     "╗",
+		BottomLeft:   "╚",
+		BottomRight:  "╝",
+		MiddleLeft:   "╠",
+		MiddleRight:  "╣",
+		Middle:       "╬",
+		MiddleTop:    "╦",
+		MiddleBottom: "╩",
+	}
+
+	hiddenBorder = Border{
+		Top:          " ",
+		Bottom:       " ",
+		Left:         " ",
+		Right:        " ",
+		TopLeft:      " ",
+		TopRight:     " ",
+		BottomLeft:   " ",
+		BottomRight:  " ",
+		MiddleLeft:   " ",
+		MiddleRight:  " ",
+		Middle:       " ",
+		MiddleTop:    " ",
+		MiddleBottom: " ",
+	}
+
+	markdownBorder = Border{
+		Top:          "-",
+		Bottom:       "-",
+		Left:         "|",
+		Right:        "|",
+		TopLeft:      "|",
+		TopRight:     "|",
+		BottomLeft:   "|",
+		BottomRight:  "|",
+		MiddleLeft:   "|",
+		MiddleRight:  "|",
+		Middle:       "|",
+		MiddleTop:    "|",
+		MiddleBottom: "|",
+	}
+
+	asciiBorder = Border{
+		Top:          "-",
+		Bottom:       "-",
+		Left:         "|",
+		Right:        "|",
+		TopLeft:      "+",
+		TopRight:     "+",
+		BottomLeft:   "+",
+		BottomRight:  "+",
+		MiddleLeft:   "+",
+		MiddleRight:  "+",
+		Middle:       "+",
+		MiddleTop:    "+",
+		MiddleBottom: "+",
+	}
+)
+
+// NormalBorder returns a standard-type border with a normal weight and 90
+// degree corners.
+func NormalBorder() Border {
+	return normalBorder
+}
+
+// RoundedBorder returns a border with rounded corners.
+func RoundedBorder() Border {
+	return roundedBorder
+}
+
+// BlockBorder returns a border that takes the whole block.
+func BlockBorder() Border {
+	return blockBorder
+}
+
+// OuterHalfBlockBorder returns a half-block border that sits outside the frame.
+func OuterHalfBlockBorder() Border {
+	return outerHalfBlockBorder
+}
+
+// InnerHalfBlockBorder returns a half-block border that sits inside the frame.
+func InnerHalfBlockBorder() Border {
+	return innerHalfBlockBorder
+}
+
+// ThickBorder returns a border that's thicker than the one returned by
+// NormalBorder.
+func ThickBorder() Border {
+	return thickBorder
+}
+
+// DoubleBorder returns a border comprised of two thin strokes.
+func DoubleBorder() Border {
+	return doubleBorder
+}
+
+// HiddenBorder returns a border that renders as a series of single-cell
+// spaces. It's useful for cases when you want to remove a standard border but
+// maintain layout positioning. This said, you can still apply a background
+// color to a hidden border.
+func HiddenBorder() Border {
+	return hiddenBorder
+}
+
+// MarkdownBorder return a table border in markdown style.
+//
+// Make sure to disable top and bottom border for the best result. This will
+// ensure that the output is valid markdown.
+//
+//	table.New().Border(lipgloss.MarkdownBorder()).BorderTop(false).BorderBottom(false)
+func MarkdownBorder() Border {
+	return markdownBorder
+}
+
+// ASCIIBorder returns a table border with ASCII characters.
+func ASCIIBorder() Border {
+	return asciiBorder
+}
+
+func (s Style) applyBorder(str string) string {
+	var (
+		border    = s.getBorderStyle()
+		hasTop    = s.getAsBool(borderTopKey, false)
+		hasRight  = s.getAsBool(borderRightKey, false)
+		hasBottom = s.getAsBool(borderBottomKey, false)
+		hasLeft   = s.getAsBool(borderLeftKey, false)
+
+		topFG    = s.getAsColor(borderTopForegroundKey)
+		rightFG  = s.getAsColor(borderRightForegroundKey)
+		bottomFG = s.getAsColor(borderBottomForegroundKey)
+		leftFG   = s.getAsColor(borderLeftForegroundKey)
+
+		topBG    = s.getAsColor(borderTopBackgroundKey)
+		rightBG  = s.getAsColor(borderRightBackgroundKey)
+		bottomBG = s.getAsColor(borderBottomBackgroundKey)
+		leftBG   = s.getAsColor(borderLeftBackgroundKey)
+	)
+
+	// If a border is set and no sides have been specifically turned on or off
+	// render borders on all sides.
+	if s.implicitBorders() {
+		hasTop = true
+		hasRight = true
+		hasBottom = true
+		hasLeft = true
+	}
+
+	// If no border is set or all borders are been disabled, abort.
+	if border == noBorder || (!hasTop && !hasRight && !hasBottom && !hasLeft) {
+		return str
+	}
+
+	lines, width := getLines(str)
+
+	if hasLeft {
+		if border.Left == "" {
+			border.Left = " "
+		}
+		width += maxRuneWidth(border.Left)
+	}
+
+	if hasRight && border.Right == "" {
+		border.Right = " "
+	}
+
+	// If corners should be rendered but are set with the empty string, fill them
+	// with a single space.
+	if hasTop && hasLeft && border.TopLeft == "" {
+		border.TopLeft = " "
+	}
+	if hasTop && hasRight && border.TopRight == "" {
+		border.TopRight = " "
+	}
+	if hasBottom && hasLeft && border.BottomLeft == "" {
+		border.BottomLeft = " "
+	}
+	if hasBottom && hasRight && border.BottomRight == "" {
+		border.BottomRight = " "
+	}
+
+	// Figure out which corners we should actually be using based on which
+	// sides are set to show.
+	if hasTop {
+		switch {
+		case !hasLeft && !hasRight:
+			border.TopLeft = ""
+			border.TopRight = ""
+		case !hasLeft:
+			border.TopLeft = ""
+		case !hasRight:
+			border.TopRight = ""
+		}
+	}
+	if hasBottom {
+		switch {
+		case !hasLeft && !hasRight:
+			border.BottomLeft = ""
+			border.BottomRight = ""
+		case !hasLeft:
+			border.BottomLeft = ""
+		case !hasRight:
+			border.BottomRight = ""
+		}
+	}
+
+	// For now, limit corners to one rune.
+	border.TopLeft = getFirstRuneAsString(border.TopLeft)
+	border.TopRight = getFirstRuneAsString(border.TopRight)
+	border.BottomRight = getFirstRuneAsString(border.BottomRight)
+	border.BottomLeft = getFirstRuneAsString(border.BottomLeft)
+
+	var out strings.Builder
+
+	// Render top
+	if hasTop {
+		top := renderHorizontalEdge(border.TopLeft, border.Top, border.TopRight, width)
+		top = s.styleBorder(top, topFG, topBG)
+		out.WriteString(top)
+		out.WriteRune('\n')
+	}
+
+	leftRunes := []rune(border.Left)
+	leftIndex := 0
+
+	rightRunes := []rune(border.Right)
+	rightIndex := 0
+
+	// Render sides
+	for i, l := range lines {
+		if hasLeft {
+			r := string(leftRunes[leftIndex])
+			leftIndex++
+			if leftIndex >= len(leftRunes) {
+				leftIndex = 0
+			}
+			out.WriteString(s.styleBorder(r, leftFG, leftBG))
+		}
+		out.WriteString(l)
+		if hasRight {
+			r := string(rightRunes[rightIndex])
+			rightIndex++
+			if rightIndex >= len(rightRunes) {
+				rightIndex = 0
+			}
+			out.WriteString(s.styleBorder(r, rightFG, rightBG))
+		}
+		if i < len(lines)-1 {
+			out.WriteRune('\n')
+		}
+	}
+
+	// Render bottom
+	if hasBottom {
+		bottom := renderHorizontalEdge(border.BottomLeft, border.Bottom, border.BottomRight, width)
+		bottom = s.styleBorder(bottom, bottomFG, bottomBG)
+		out.WriteRune('\n')
+		out.WriteString(bottom)
+	}
+
+	return out.String()
+}
+
+// Render the horizontal (top or bottom) portion of a border.
+func renderHorizontalEdge(left, middle, right string, width int) string {
+	if middle == "" {
+		middle = " "
+	}
+
+	leftWidth := ansi.StringWidth(left)
+	rightWidth := ansi.StringWidth(right)
+
+	runes := []rune(middle)
+	j := 0
+
+	out := strings.Builder{}
+	out.WriteString(left)
+	for i := leftWidth + rightWidth; i < width+rightWidth; {
+		out.WriteRune(runes[j])
+		j++
+		if j >= len(runes) {
+			j = 0
+		}
+		i += ansi.StringWidth(string(runes[j]))
+	}
+	out.WriteString(right)
+
+	return out.String()
+}
+
+// Apply foreground and background styling to a border.
+func (s Style) styleBorder(border string, fg, bg TerminalColor) string {
+	if fg == noColor && bg == noColor {
+		return border
+	}
+
+	style := termenv.Style{}
+
+	if fg != noColor {
+		style = style.Foreground(fg.color(s.r))
+	}
+	if bg != noColor {
+		style = style.Background(bg.color(s.r))
+	}
+
+	return style.Styled(border)
+}
+
+func maxRuneWidth(str string) int {
+	var width int
+
+	state := -1
+	for len(str) > 0 {
+		var w int
+		_, str, w, state = uniseg.FirstGraphemeClusterInString(str, state)
+		if w > width {
+			width = w
+		}
+	}
+
+	return width
+}
+
+func getFirstRuneAsString(str string) string {
+	if str == "" {
+		return str
+	}
+	r := []rune(str)
+	return string(r[0])
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/color.go b/vendor/github.com/charmbracelet/lipgloss/color.go
new file mode 100644
index 0000000..6caf3a3
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/color.go
@@ -0,0 +1,172 @@
+package lipgloss
+
+import (
+	"strconv"
+
+	"github.com/muesli/termenv"
+)
+
+// TerminalColor is a color intended to be rendered in the terminal.
+type TerminalColor interface {
+	color(*Renderer) termenv.Color
+	RGBA() (r, g, b, a uint32)
+}
+
+var noColor = NoColor{}
+
+// NoColor is used to specify the absence of color styling. When this is active
+// foreground colors will be rendered with the terminal's default text color,
+// and background colors will not be drawn at all.
+//
+// Example usage:
+//
+//	var style = someStyle.Background(lipgloss.NoColor{})
+type NoColor struct{}
+
+func (NoColor) color(*Renderer) termenv.Color {
+	return termenv.NoColor{}
+}
+
+// RGBA returns the RGBA value of this color. Because we have to return
+// something, despite this color being the absence of color, we're returning
+// black with 100% opacity.
+//
+// Red: 0x0, Green: 0x0, Blue: 0x0, Alpha: 0xFFFF.
+//
+// Deprecated.
+func (n NoColor) RGBA() (r, g, b, a uint32) {
+	return 0x0, 0x0, 0x0, 0xFFFF //nolint:mnd
+}
+
+// Color specifies a color by hex or ANSI value. For example:
+//
+//	ansiColor := lipgloss.Color("21")
+//	hexColor := lipgloss.Color("#0000ff")
+type Color string
+
+func (c Color) color(r *Renderer) termenv.Color {
+	return r.ColorProfile().Color(string(c))
+}
+
+// RGBA returns the RGBA value of this color. This satisfies the Go Color
+// interface. Note that on error we return black with 100% opacity, or:
+//
+// Red: 0x0, Green: 0x0, Blue: 0x0, Alpha: 0xFFFF.
+//
+// Deprecated.
+func (c Color) RGBA() (r, g, b, a uint32) {
+	return termenv.ConvertToRGB(c.color(renderer)).RGBA()
+}
+
+// ANSIColor is a color specified by an ANSI color value. It's merely syntactic
+// sugar for the more general Color function. Invalid colors will render as
+// black.
+//
+// Example usage:
+//
+//	// These two statements are equivalent.
+//	colorA := lipgloss.ANSIColor(21)
+//	colorB := lipgloss.Color("21")
+type ANSIColor uint
+
+func (ac ANSIColor) color(r *Renderer) termenv.Color {
+	return Color(strconv.FormatUint(uint64(ac), 10)).color(r)
+}
+
+// RGBA returns the RGBA value of this color. This satisfies the Go Color
+// interface. Note that on error we return black with 100% opacity, or:
+//
+// Red: 0x0, Green: 0x0, Blue: 0x0, Alpha: 0xFFFF.
+//
+// Deprecated.
+func (ac ANSIColor) RGBA() (r, g, b, a uint32) {
+	cf := Color(strconv.FormatUint(uint64(ac), 10))
+	return cf.RGBA()
+}
+
+// AdaptiveColor provides color options for light and dark backgrounds. The
+// appropriate color will be returned at runtime based on the darkness of the
+// terminal background color.
+//
+// Example usage:
+//
+//	color := lipgloss.AdaptiveColor{Light: "#0000ff", Dark: "#000099"}
+type AdaptiveColor struct {
+	Light string
+	Dark  string
+}
+
+func (ac AdaptiveColor) color(r *Renderer) termenv.Color {
+	if r.HasDarkBackground() {
+		return Color(ac.Dark).color(r)
+	}
+	return Color(ac.Light).color(r)
+}
+
+// RGBA returns the RGBA value of this color. This satisfies the Go Color
+// interface. Note that on error we return black with 100% opacity, or:
+//
+// Red: 0x0, Green: 0x0, Blue: 0x0, Alpha: 0xFFFF.
+//
+// Deprecated.
+func (ac AdaptiveColor) RGBA() (r, g, b, a uint32) {
+	return termenv.ConvertToRGB(ac.color(renderer)).RGBA()
+}
+
+// CompleteColor specifies exact values for truecolor, ANSI256, and ANSI color
+// profiles. Automatic color degradation will not be performed.
+type CompleteColor struct {
+	TrueColor string
+	ANSI256   string
+	ANSI      string
+}
+
+func (c CompleteColor) color(r *Renderer) termenv.Color {
+	p := r.ColorProfile()
+	switch p { //nolint:exhaustive
+	case termenv.TrueColor:
+		return p.Color(c.TrueColor)
+	case termenv.ANSI256:
+		return p.Color(c.ANSI256)
+	case termenv.ANSI:
+		return p.Color(c.ANSI)
+	default:
+		return termenv.NoColor{}
+	}
+}
+
+// RGBA returns the RGBA value of this color. This satisfies the Go Color
+// interface. Note that on error we return black with 100% opacity, or:
+//
+// Red: 0x0, Green: 0x0, Blue: 0x0, Alpha: 0xFFFF.
+// CompleteAdaptiveColor specifies exact values for truecolor, ANSI256, and ANSI color
+//
+// Deprecated.
+func (c CompleteColor) RGBA() (r, g, b, a uint32) {
+	return termenv.ConvertToRGB(c.color(renderer)).RGBA()
+}
+
+// CompleteAdaptiveColor specifies exact values for truecolor, ANSI256, and ANSI color
+// profiles, with separate options for light and dark backgrounds. Automatic
+// color degradation will not be performed.
+type CompleteAdaptiveColor struct {
+	Light CompleteColor
+	Dark  CompleteColor
+}
+
+func (cac CompleteAdaptiveColor) color(r *Renderer) termenv.Color {
+	if r.HasDarkBackground() {
+		return cac.Dark.color(r)
+	}
+	return cac.Light.color(r)
+}
+
+// RGBA returns the RGBA value of this color. This satisfies the Go Color
+// interface. Note that on error we return black with 100% opacity, or:
+//
+// Red: 0x0, Green: 0x0, Blue: 0x0, Alpha: 0xFFFF.
+//
+// Deprecated.
+func (cac CompleteAdaptiveColor) RGBA() (r, g, b, a uint32) {
+	return termenv.ConvertToRGB(cac.color(renderer)).RGBA()
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/get.go b/vendor/github.com/charmbracelet/lipgloss/get.go
new file mode 100644
index 0000000..422b4ce
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/get.go
@@ -0,0 +1,556 @@
+package lipgloss
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// GetBold returns the style's bold value. If no value is set false is returned.
+func (s Style) GetBold() bool {
+	return s.getAsBool(boldKey, false)
+}
+
+// GetItalic returns the style's italic value. If no value is set false is
+// returned.
+func (s Style) GetItalic() bool {
+	return s.getAsBool(italicKey, false)
+}
+
+// GetUnderline returns the style's underline value. If no value is set false is
+// returned.
+func (s Style) GetUnderline() bool {
+	return s.getAsBool(underlineKey, false)
+}
+
+// GetStrikethrough returns the style's strikethrough value. If no value is set false
+// is returned.
+func (s Style) GetStrikethrough() bool {
+	return s.getAsBool(strikethroughKey, false)
+}
+
+// GetReverse returns the style's reverse value. If no value is set false is
+// returned.
+func (s Style) GetReverse() bool {
+	return s.getAsBool(reverseKey, false)
+}
+
+// GetBlink returns the style's blink value. If no value is set false is
+// returned.
+func (s Style) GetBlink() bool {
+	return s.getAsBool(blinkKey, false)
+}
+
+// GetFaint returns the style's faint value. If no value is set false is
+// returned.
+func (s Style) GetFaint() bool {
+	return s.getAsBool(faintKey, false)
+}
+
+// GetForeground returns the style's foreground color. If no value is set
+// NoColor{} is returned.
+func (s Style) GetForeground() TerminalColor {
+	return s.getAsColor(foregroundKey)
+}
+
+// GetBackground returns the style's background color. If no value is set
+// NoColor{} is returned.
+func (s Style) GetBackground() TerminalColor {
+	return s.getAsColor(backgroundKey)
+}
+
+// GetWidth returns the style's width setting. If no width is set 0 is
+// returned.
+func (s Style) GetWidth() int {
+	return s.getAsInt(widthKey)
+}
+
+// GetHeight returns the style's height setting. If no height is set 0 is
+// returned.
+func (s Style) GetHeight() int {
+	return s.getAsInt(heightKey)
+}
+
+// GetAlign returns the style's implicit horizontal alignment setting.
+// If no alignment is set Position.Left is returned.
+func (s Style) GetAlign() Position {
+	v := s.getAsPosition(alignHorizontalKey)
+	if v == Position(0) {
+		return Left
+	}
+	return v
+}
+
+// GetAlignHorizontal returns the style's implicit horizontal alignment setting.
+// If no alignment is set Position.Left is returned.
+func (s Style) GetAlignHorizontal() Position {
+	v := s.getAsPosition(alignHorizontalKey)
+	if v == Position(0) {
+		return Left
+	}
+	return v
+}
+
+// GetAlignVertical returns the style's implicit vertical alignment setting.
+// If no alignment is set Position.Top is returned.
+func (s Style) GetAlignVertical() Position {
+	v := s.getAsPosition(alignVerticalKey)
+	if v == Position(0) {
+		return Top
+	}
+	return v
+}
+
+// GetPadding returns the style's top, right, bottom, and left padding values,
+// in that order. 0 is returned for unset values.
+func (s Style) GetPadding() (top, right, bottom, left int) {
+	return s.getAsInt(paddingTopKey),
+		s.getAsInt(paddingRightKey),
+		s.getAsInt(paddingBottomKey),
+		s.getAsInt(paddingLeftKey)
+}
+
+// GetPaddingTop returns the style's top padding. If no value is set 0 is
+// returned.
+func (s Style) GetPaddingTop() int {
+	return s.getAsInt(paddingTopKey)
+}
+
+// GetPaddingRight returns the style's right padding. If no value is set 0 is
+// returned.
+func (s Style) GetPaddingRight() int {
+	return s.getAsInt(paddingRightKey)
+}
+
+// GetPaddingBottom returns the style's bottom padding. If no value is set 0 is
+// returned.
+func (s Style) GetPaddingBottom() int {
+	return s.getAsInt(paddingBottomKey)
+}
+
+// GetPaddingLeft returns the style's left padding. If no value is set 0 is
+// returned.
+func (s Style) GetPaddingLeft() int {
+	return s.getAsInt(paddingLeftKey)
+}
+
+// GetHorizontalPadding returns the style's left and right padding. Unset
+// values are measured as 0.
+func (s Style) GetHorizontalPadding() int {
+	return s.getAsInt(paddingLeftKey) + s.getAsInt(paddingRightKey)
+}
+
+// GetVerticalPadding returns the style's top and bottom padding. Unset values
+// are measured as 0.
+func (s Style) GetVerticalPadding() int {
+	return s.getAsInt(paddingTopKey) + s.getAsInt(paddingBottomKey)
+}
+
+// GetColorWhitespace returns the style's whitespace coloring setting. If no
+// value is set false is returned.
+func (s Style) GetColorWhitespace() bool {
+	return s.getAsBool(colorWhitespaceKey, false)
+}
+
+// GetMargin returns the style's top, right, bottom, and left margins, in that
+// order. 0 is returned for unset values.
+func (s Style) GetMargin() (top, right, bottom, left int) {
+	return s.getAsInt(marginTopKey),
+		s.getAsInt(marginRightKey),
+		s.getAsInt(marginBottomKey),
+		s.getAsInt(marginLeftKey)
+}
+
+// GetMarginTop returns the style's top margin. If no value is set 0 is
+// returned.
+func (s Style) GetMarginTop() int {
+	return s.getAsInt(marginTopKey)
+}
+
+// GetMarginRight returns the style's right margin. If no value is set 0 is
+// returned.
+func (s Style) GetMarginRight() int {
+	return s.getAsInt(marginRightKey)
+}
+
+// GetMarginBottom returns the style's bottom margin. If no value is set 0 is
+// returned.
+func (s Style) GetMarginBottom() int {
+	return s.getAsInt(marginBottomKey)
+}
+
+// GetMarginLeft returns the style's left margin. If no value is set 0 is
+// returned.
+func (s Style) GetMarginLeft() int {
+	return s.getAsInt(marginLeftKey)
+}
+
+// GetHorizontalMargins returns the style's left and right margins. Unset
+// values are measured as 0.
+func (s Style) GetHorizontalMargins() int {
+	return s.getAsInt(marginLeftKey) + s.getAsInt(marginRightKey)
+}
+
+// GetVerticalMargins returns the style's top and bottom margins. Unset values
+// are measured as 0.
+func (s Style) GetVerticalMargins() int {
+	return s.getAsInt(marginTopKey) + s.getAsInt(marginBottomKey)
+}
+
+// GetBorder returns the style's border style (type Border) and value for the
+// top, right, bottom, and left in that order. If no value is set for the
+// border style, Border{} is returned. For all other unset values false is
+// returned.
+func (s Style) GetBorder() (b Border, top, right, bottom, left bool) {
+	return s.getBorderStyle(),
+		s.getAsBool(borderTopKey, false),
+		s.getAsBool(borderRightKey, false),
+		s.getAsBool(borderBottomKey, false),
+		s.getAsBool(borderLeftKey, false)
+}
+
+// GetBorderStyle returns the style's border style (type Border). If no value
+// is set Border{} is returned.
+func (s Style) GetBorderStyle() Border {
+	return s.getBorderStyle()
+}
+
+// GetBorderTop returns the style's top border setting. If no value is set
+// false is returned.
+func (s Style) GetBorderTop() bool {
+	return s.getAsBool(borderTopKey, false)
+}
+
+// GetBorderRight returns the style's right border setting. If no value is set
+// false is returned.
+func (s Style) GetBorderRight() bool {
+	return s.getAsBool(borderRightKey, false)
+}
+
+// GetBorderBottom returns the style's bottom border setting. If no value is
+// set false is returned.
+func (s Style) GetBorderBottom() bool {
+	return s.getAsBool(borderBottomKey, false)
+}
+
+// GetBorderLeft returns the style's left border setting. If no value is
+// set false is returned.
+func (s Style) GetBorderLeft() bool {
+	return s.getAsBool(borderLeftKey, false)
+}
+
+// GetBorderTopForeground returns the style's border top foreground color. If
+// no value is set NoColor{} is returned.
+func (s Style) GetBorderTopForeground() TerminalColor {
+	return s.getAsColor(borderTopForegroundKey)
+}
+
+// GetBorderRightForeground returns the style's border right foreground color.
+// If no value is set NoColor{} is returned.
+func (s Style) GetBorderRightForeground() TerminalColor {
+	return s.getAsColor(borderRightForegroundKey)
+}
+
+// GetBorderBottomForeground returns the style's border bottom foreground
+// color.  If no value is set NoColor{} is returned.
+func (s Style) GetBorderBottomForeground() TerminalColor {
+	return s.getAsColor(borderBottomForegroundKey)
+}
+
+// GetBorderLeftForeground returns the style's border left foreground
+// color.  If no value is set NoColor{} is returned.
+func (s Style) GetBorderLeftForeground() TerminalColor {
+	return s.getAsColor(borderLeftForegroundKey)
+}
+
+// GetBorderTopBackground returns the style's border top background color. If
+// no value is set NoColor{} is returned.
+func (s Style) GetBorderTopBackground() TerminalColor {
+	return s.getAsColor(borderTopBackgroundKey)
+}
+
+// GetBorderRightBackground returns the style's border right background color.
+// If no value is set NoColor{} is returned.
+func (s Style) GetBorderRightBackground() TerminalColor {
+	return s.getAsColor(borderRightBackgroundKey)
+}
+
+// GetBorderBottomBackground returns the style's border bottom background
+// color.  If no value is set NoColor{} is returned.
+func (s Style) GetBorderBottomBackground() TerminalColor {
+	return s.getAsColor(borderBottomBackgroundKey)
+}
+
+// GetBorderLeftBackground returns the style's border left background
+// color.  If no value is set NoColor{} is returned.
+func (s Style) GetBorderLeftBackground() TerminalColor {
+	return s.getAsColor(borderLeftBackgroundKey)
+}
+
+// GetBorderTopWidth returns the width of the top border. If borders contain
+// runes of varying widths, the widest rune is returned. If no border exists on
+// the top edge, 0 is returned.
+//
+// Deprecated: This function simply calls Style.GetBorderTopSize.
+func (s Style) GetBorderTopWidth() int {
+	return s.GetBorderTopSize()
+}
+
+// GetBorderTopSize returns the width of the top border. If borders contain
+// runes of varying widths, the widest rune is returned. If no border exists on
+// the top edge, 0 is returned.
+func (s Style) GetBorderTopSize() int {
+	if !s.getAsBool(borderTopKey, false) && !s.implicitBorders() {
+		return 0
+	}
+	return s.getBorderStyle().GetTopSize()
+}
+
+// GetBorderLeftSize returns the width of the left border. If borders contain
+// runes of varying widths, the widest rune is returned. If no border exists on
+// the left edge, 0 is returned.
+func (s Style) GetBorderLeftSize() int {
+	if !s.getAsBool(borderLeftKey, false) && !s.implicitBorders() {
+		return 0
+	}
+	return s.getBorderStyle().GetLeftSize()
+}
+
+// GetBorderBottomSize returns the width of the bottom border. If borders
+// contain runes of varying widths, the widest rune is returned. If no border
+// exists on the left edge, 0 is returned.
+func (s Style) GetBorderBottomSize() int {
+	if !s.getAsBool(borderBottomKey, false) && !s.implicitBorders() {
+		return 0
+	}
+	return s.getBorderStyle().GetBottomSize()
+}
+
+// GetBorderRightSize returns the width of the right border. If borders
+// contain runes of varying widths, the widest rune is returned. If no border
+// exists on the right edge, 0 is returned.
+func (s Style) GetBorderRightSize() int {
+	if !s.getAsBool(borderRightKey, false) && !s.implicitBorders() {
+		return 0
+	}
+	return s.getBorderStyle().GetRightSize()
+}
+
+// GetHorizontalBorderSize returns the width of the horizontal borders. If
+// borders contain runes of varying widths, the widest rune is returned. If no
+// border exists on the horizontal edges, 0 is returned.
+func (s Style) GetHorizontalBorderSize() int {
+	return s.GetBorderLeftSize() + s.GetBorderRightSize()
+}
+
+// GetVerticalBorderSize returns the width of the vertical borders. If
+// borders contain runes of varying widths, the widest rune is returned. If no
+// border exists on the vertical edges, 0 is returned.
+func (s Style) GetVerticalBorderSize() int {
+	return s.GetBorderTopSize() + s.GetBorderBottomSize()
+}
+
+// GetInline returns the style's inline setting. If no value is set false is
+// returned.
+func (s Style) GetInline() bool {
+	return s.getAsBool(inlineKey, false)
+}
+
+// GetMaxWidth returns the style's max width setting. If no value is set 0 is
+// returned.
+func (s Style) GetMaxWidth() int {
+	return s.getAsInt(maxWidthKey)
+}
+
+// GetMaxHeight returns the style's max height setting. If no value is set 0 is
+// returned.
+func (s Style) GetMaxHeight() int {
+	return s.getAsInt(maxHeightKey)
+}
+
+// GetTabWidth returns the style's tab width setting. If no value is set 4 is
+// returned which is the implicit default.
+func (s Style) GetTabWidth() int {
+	return s.getAsInt(tabWidthKey)
+}
+
+// GetUnderlineSpaces returns whether or not the style is set to underline
+// spaces. If not value is set false is returned.
+func (s Style) GetUnderlineSpaces() bool {
+	return s.getAsBool(underlineSpacesKey, false)
+}
+
+// GetStrikethroughSpaces returns whether or not the style is set to strikethrough
+// spaces. If not value is set false is returned.
+func (s Style) GetStrikethroughSpaces() bool {
+	return s.getAsBool(strikethroughSpacesKey, false)
+}
+
+// GetHorizontalFrameSize returns the sum of the style's horizontal margins, padding
+// and border widths.
+//
+// Provisional: this method may be renamed.
+func (s Style) GetHorizontalFrameSize() int {
+	return s.GetHorizontalMargins() + s.GetHorizontalPadding() + s.GetHorizontalBorderSize()
+}
+
+// GetVerticalFrameSize returns the sum of the style's vertical margins, padding
+// and border widths.
+//
+// Provisional: this method may be renamed.
+func (s Style) GetVerticalFrameSize() int {
+	return s.GetVerticalMargins() + s.GetVerticalPadding() + s.GetVerticalBorderSize()
+}
+
+// GetFrameSize returns the sum of the margins, padding and border width for
+// both the horizontal and vertical margins.
+func (s Style) GetFrameSize() (x, y int) {
+	return s.GetHorizontalFrameSize(), s.GetVerticalFrameSize()
+}
+
+// GetTransform returns the transform set on the style. If no transform is set
+// nil is returned.
+func (s Style) GetTransform() func(string) string {
+	return s.getAsTransform(transformKey)
+}
+
+// Returns whether or not the given property is set.
+func (s Style) isSet(k propKey) bool {
+	return s.props.has(k)
+}
+
+func (s Style) getAsBool(k propKey, defaultVal bool) bool {
+	if !s.isSet(k) {
+		return defaultVal
+	}
+	return s.attrs&int(k) != 0
+}
+
+func (s Style) getAsColor(k propKey) TerminalColor {
+	if !s.isSet(k) {
+		return noColor
+	}
+
+	var c TerminalColor
+	switch k { //nolint:exhaustive
+	case foregroundKey:
+		c = s.fgColor
+	case backgroundKey:
+		c = s.bgColor
+	case marginBackgroundKey:
+		c = s.marginBgColor
+	case borderTopForegroundKey:
+		c = s.borderTopFgColor
+	case borderRightForegroundKey:
+		c = s.borderRightFgColor
+	case borderBottomForegroundKey:
+		c = s.borderBottomFgColor
+	case borderLeftForegroundKey:
+		c = s.borderLeftFgColor
+	case borderTopBackgroundKey:
+		c = s.borderTopBgColor
+	case borderRightBackgroundKey:
+		c = s.borderRightBgColor
+	case borderBottomBackgroundKey:
+		c = s.borderBottomBgColor
+	case borderLeftBackgroundKey:
+		c = s.borderLeftBgColor
+	}
+
+	if c != nil {
+		return c
+	}
+
+	return noColor
+}
+
+func (s Style) getAsInt(k propKey) int {
+	if !s.isSet(k) {
+		return 0
+	}
+	switch k { //nolint:exhaustive
+	case widthKey:
+		return s.width
+	case heightKey:
+		return s.height
+	case paddingTopKey:
+		return s.paddingTop
+	case paddingRightKey:
+		return s.paddingRight
+	case paddingBottomKey:
+		return s.paddingBottom
+	case paddingLeftKey:
+		return s.paddingLeft
+	case marginTopKey:
+		return s.marginTop
+	case marginRightKey:
+		return s.marginRight
+	case marginBottomKey:
+		return s.marginBottom
+	case marginLeftKey:
+		return s.marginLeft
+	case maxWidthKey:
+		return s.maxWidth
+	case maxHeightKey:
+		return s.maxHeight
+	case tabWidthKey:
+		return s.tabWidth
+	}
+	return 0
+}
+
+func (s Style) getAsPosition(k propKey) Position {
+	if !s.isSet(k) {
+		return Position(0)
+	}
+	switch k { //nolint:exhaustive
+	case alignHorizontalKey:
+		return s.alignHorizontal
+	case alignVerticalKey:
+		return s.alignVertical
+	}
+	return Position(0)
+}
+
+func (s Style) getBorderStyle() Border {
+	if !s.isSet(borderStyleKey) {
+		return noBorder
+	}
+	return s.borderStyle
+}
+
+// Returns whether or not the style has implicit borders. This happens when
+// a border style has been set but no border sides have been explicitly turned
+// on or off.
+func (s Style) implicitBorders() bool {
+	var (
+		borderStyle = s.getBorderStyle()
+		topSet      = s.isSet(borderTopKey)
+		rightSet    = s.isSet(borderRightKey)
+		bottomSet   = s.isSet(borderBottomKey)
+		leftSet     = s.isSet(borderLeftKey)
+	)
+	return borderStyle != noBorder && !(topSet || rightSet || bottomSet || leftSet)
+}
+
+func (s Style) getAsTransform(propKey) func(string) string {
+	if !s.isSet(transformKey) {
+		return nil
+	}
+	return s.transform
+}
+
+// Split a string into lines, additionally returning the size of the widest
+// line.
+func getLines(s string) (lines []string, widest int) {
+	lines = strings.Split(s, "\n")
+
+	for _, l := range lines {
+		w := ansi.StringWidth(l)
+		if widest < w {
+			widest = w
+		}
+	}
+
+	return lines, widest
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/join.go b/vendor/github.com/charmbracelet/lipgloss/join.go
new file mode 100644
index 0000000..b0a23a5
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/join.go
@@ -0,0 +1,175 @@
+package lipgloss
+
+import (
+	"math"
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// JoinHorizontal is a utility function for horizontally joining two
+// potentially multi-lined strings along a vertical axis. The first argument is
+// the position, with 0 being all the way at the top and 1 being all the way
+// at the bottom.
+//
+// If you just want to align to the top, center or bottom you may as well just
+// use the helper constants Top, Center, and Bottom.
+//
+// Example:
+//
+//	blockB := "...\n...\n..."
+//	blockA := "...\n...\n...\n...\n..."
+//
+//	// Join 20% from the top
+//	str := lipgloss.JoinHorizontal(0.2, blockA, blockB)
+//
+//	// Join on the top edge
+//	str := lipgloss.JoinHorizontal(lipgloss.Top, blockA, blockB)
+func JoinHorizontal(pos Position, strs ...string) string {
+	if len(strs) == 0 {
+		return ""
+	}
+	if len(strs) == 1 {
+		return strs[0]
+	}
+
+	var (
+		// Groups of strings broken into multiple lines
+		blocks = make([][]string, len(strs))
+
+		// Max line widths for the above text blocks
+		maxWidths = make([]int, len(strs))
+
+		// Height of the tallest block
+		maxHeight int
+	)
+
+	// Break text blocks into lines and get max widths for each text block
+	for i, str := range strs {
+		blocks[i], maxWidths[i] = getLines(str)
+		if len(blocks[i]) > maxHeight {
+			maxHeight = len(blocks[i])
+		}
+	}
+
+	// Add extra lines to make each side the same height
+	for i := range blocks {
+		if len(blocks[i]) >= maxHeight {
+			continue
+		}
+
+		extraLines := make([]string, maxHeight-len(blocks[i]))
+
+		switch pos { //nolint:exhaustive
+		case Top:
+			blocks[i] = append(blocks[i], extraLines...)
+
+		case Bottom:
+			blocks[i] = append(extraLines, blocks[i]...)
+
+		default: // Somewhere in the middle
+			n := len(extraLines)
+			split := int(math.Round(float64(n) * pos.value()))
+			top := n - split
+			bottom := n - top
+
+			blocks[i] = append(extraLines[top:], blocks[i]...)
+			blocks[i] = append(blocks[i], extraLines[bottom:]...)
+		}
+	}
+
+	// Merge lines
+	var b strings.Builder
+	for i := range blocks[0] { // remember, all blocks have the same number of members now
+		for j, block := range blocks {
+			b.WriteString(block[i])
+
+			// Also make lines the same length
+			b.WriteString(strings.Repeat(" ", maxWidths[j]-ansi.StringWidth(block[i])))
+		}
+		if i < len(blocks[0])-1 {
+			b.WriteRune('\n')
+		}
+	}
+
+	return b.String()
+}
+
+// JoinVertical is a utility function for vertically joining two potentially
+// multi-lined strings along a horizontal axis. The first argument is the
+// position, with 0 being all the way to the left and 1 being all the way to
+// the right.
+//
+// If you just want to align to the left, right or center you may as well just
+// use the helper constants Left, Center, and Right.
+//
+// Example:
+//
+//	blockB := "...\n...\n..."
+//	blockA := "...\n...\n...\n...\n..."
+//
+//	// Join 20% from the top
+//	str := lipgloss.JoinVertical(0.2, blockA, blockB)
+//
+//	// Join on the right edge
+//	str := lipgloss.JoinVertical(lipgloss.Right, blockA, blockB)
+func JoinVertical(pos Position, strs ...string) string {
+	if len(strs) == 0 {
+		return ""
+	}
+	if len(strs) == 1 {
+		return strs[0]
+	}
+
+	var (
+		blocks   = make([][]string, len(strs))
+		maxWidth int
+	)
+
+	for i := range strs {
+		var w int
+		blocks[i], w = getLines(strs[i])
+		if w > maxWidth {
+			maxWidth = w
+		}
+	}
+
+	var b strings.Builder
+	for i, block := range blocks {
+		for j, line := range block {
+			w := maxWidth - ansi.StringWidth(line)
+
+			switch pos { //nolint:exhaustive
+			case Left:
+				b.WriteString(line)
+				b.WriteString(strings.Repeat(" ", w))
+
+			case Right:
+				b.WriteString(strings.Repeat(" ", w))
+				b.WriteString(line)
+
+			default: // Somewhere in the middle
+				if w < 1 {
+					b.WriteString(line)
+					break
+				}
+
+				split := int(math.Round(float64(w) * pos.value()))
+				right := w - split
+				left := w - right
+
+				b.WriteString(strings.Repeat(" ", left))
+				b.WriteString(line)
+				b.WriteString(strings.Repeat(" ", right))
+			}
+
+			// Write a newline as long as we're not on the last line of the
+			// last block.
+			if !(i == len(blocks)-1 && j == len(block)-1) {
+				b.WriteRune('\n')
+			}
+		}
+	}
+
+	return b.String()
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/position.go b/vendor/github.com/charmbracelet/lipgloss/position.go
new file mode 100644
index 0000000..185f5af
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/position.go
@@ -0,0 +1,154 @@
+package lipgloss
+
+import (
+	"math"
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// Position represents a position along a horizontal or vertical axis. It's in
+// situations where an axis is involved, like alignment, joining, placement and
+// so on.
+//
+// A value of 0 represents the start (the left or top) and 1 represents the end
+// (the right or bottom). 0.5 represents the center.
+//
+// There are constants Top, Bottom, Center, Left and Right in this package that
+// can be used to aid readability.
+type Position float64
+
+func (p Position) value() float64 {
+	return math.Min(1, math.Max(0, float64(p)))
+}
+
+// Position aliases.
+const (
+	Top    Position = 0.0
+	Bottom Position = 1.0
+	Center Position = 0.5
+	Left   Position = 0.0
+	Right  Position = 1.0
+)
+
+// Place places a string or text block vertically in an unstyled box of a given
+// width or height.
+func Place(width, height int, hPos, vPos Position, str string, opts ...WhitespaceOption) string {
+	return renderer.Place(width, height, hPos, vPos, str, opts...)
+}
+
+// Place places a string or text block vertically in an unstyled box of a given
+// width or height.
+func (r *Renderer) Place(width, height int, hPos, vPos Position, str string, opts ...WhitespaceOption) string {
+	return r.PlaceVertical(height, vPos, r.PlaceHorizontal(width, hPos, str, opts...), opts...)
+}
+
+// PlaceHorizontal places a string or text block horizontally in an unstyled
+// block of a given width. If the given width is shorter than the max width of
+// the string (measured by its longest line) this will be a noop.
+func PlaceHorizontal(width int, pos Position, str string, opts ...WhitespaceOption) string {
+	return renderer.PlaceHorizontal(width, pos, str, opts...)
+}
+
+// PlaceHorizontal places a string or text block horizontally in an unstyled
+// block of a given width. If the given width is shorter than the max width of
+// the string (measured by its longest line) this will be a noöp.
+func (r *Renderer) PlaceHorizontal(width int, pos Position, str string, opts ...WhitespaceOption) string {
+	lines, contentWidth := getLines(str)
+	gap := width - contentWidth
+
+	if gap <= 0 {
+		return str
+	}
+
+	ws := newWhitespace(r, opts...)
+
+	var b strings.Builder
+	for i, l := range lines {
+		// Is this line shorter than the longest line?
+		short := max(0, contentWidth-ansi.StringWidth(l))
+
+		switch pos { //nolint:exhaustive
+		case Left:
+			b.WriteString(l)
+			b.WriteString(ws.render(gap + short))
+
+		case Right:
+			b.WriteString(ws.render(gap + short))
+			b.WriteString(l)
+
+		default: // somewhere in the middle
+			totalGap := gap + short
+
+			split := int(math.Round(float64(totalGap) * pos.value()))
+			left := totalGap - split
+			right := totalGap - left
+
+			b.WriteString(ws.render(left))
+			b.WriteString(l)
+			b.WriteString(ws.render(right))
+		}
+
+		if i < len(lines)-1 {
+			b.WriteRune('\n')
+		}
+	}
+
+	return b.String()
+}
+
+// PlaceVertical places a string or text block vertically in an unstyled block
+// of a given height. If the given height is shorter than the height of the
+// string (measured by its newlines) then this will be a noop.
+func PlaceVertical(height int, pos Position, str string, opts ...WhitespaceOption) string {
+	return renderer.PlaceVertical(height, pos, str, opts...)
+}
+
+// PlaceVertical places a string or text block vertically in an unstyled block
+// of a given height. If the given height is shorter than the height of the
+// string (measured by its newlines) then this will be a noöp.
+func (r *Renderer) PlaceVertical(height int, pos Position, str string, opts ...WhitespaceOption) string {
+	contentHeight := strings.Count(str, "\n") + 1
+	gap := height - contentHeight
+
+	if gap <= 0 {
+		return str
+	}
+
+	ws := newWhitespace(r, opts...)
+
+	_, width := getLines(str)
+	emptyLine := ws.render(width)
+	b := strings.Builder{}
+
+	switch pos { //nolint:exhaustive
+	case Top:
+		b.WriteString(str)
+		b.WriteRune('\n')
+		for i := 0; i < gap; i++ {
+			b.WriteString(emptyLine)
+			if i < gap-1 {
+				b.WriteRune('\n')
+			}
+		}
+
+	case Bottom:
+		b.WriteString(strings.Repeat(emptyLine+"\n", gap))
+		b.WriteString(str)
+
+	default: // Somewhere in the middle
+		split := int(math.Round(float64(gap) * pos.value()))
+		top := gap - split
+		bottom := gap - top
+
+		b.WriteString(strings.Repeat(emptyLine+"\n", top))
+		b.WriteString(str)
+
+		for i := 0; i < bottom; i++ {
+			b.WriteRune('\n')
+			b.WriteString(emptyLine)
+		}
+	}
+
+	return b.String()
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/ranges.go b/vendor/github.com/charmbracelet/lipgloss/ranges.go
new file mode 100644
index 0000000..d171699
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/ranges.go
@@ -0,0 +1,48 @@
+package lipgloss
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// StyleRanges allows to, given a string, style ranges of it differently.
+// The function will take into account existing styles.
+// Ranges should not overlap.
+func StyleRanges(s string, ranges ...Range) string {
+	if len(ranges) == 0 {
+		return s
+	}
+
+	var buf strings.Builder
+	lastIdx := 0
+	stripped := ansi.Strip(s)
+
+	// Use Truncate and TruncateLeft to style match.MatchedIndexes without
+	// losing the original option style:
+	for _, rng := range ranges {
+		// Add the text before this match
+		if rng.Start > lastIdx {
+			buf.WriteString(ansi.Cut(s, lastIdx, rng.Start))
+		}
+		// Add the matched range with its highlight
+		buf.WriteString(rng.Style.Render(ansi.Cut(stripped, rng.Start, rng.End)))
+		lastIdx = rng.End
+	}
+
+	// Add any remaining text after the last match
+	buf.WriteString(ansi.TruncateLeft(s, lastIdx, ""))
+
+	return buf.String()
+}
+
+// NewRange returns a range that can be used with [StyleRanges].
+func NewRange(start, end int, style Style) Range {
+	return Range{start, end, style}
+}
+
+// Range to be used with [StyleRanges].
+type Range struct {
+	Start, End int
+	Style      Style
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/renderer.go b/vendor/github.com/charmbracelet/lipgloss/renderer.go
new file mode 100644
index 0000000..233aa7c
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/renderer.go
@@ -0,0 +1,181 @@
+package lipgloss
+
+import (
+	"io"
+	"sync"
+
+	"github.com/muesli/termenv"
+)
+
+// We're manually creating the struct here to avoid initializing the output and
+// query the terminal multiple times.
+var renderer = &Renderer{
+	output: termenv.DefaultOutput(),
+}
+
+// Renderer is a lipgloss terminal renderer.
+type Renderer struct {
+	output            *termenv.Output
+	colorProfile      termenv.Profile
+	hasDarkBackground bool
+
+	getColorProfile      sync.Once
+	explicitColorProfile bool
+
+	getBackgroundColor      sync.Once
+	explicitBackgroundColor bool
+
+	mtx sync.RWMutex
+}
+
+// DefaultRenderer returns the default renderer.
+func DefaultRenderer() *Renderer {
+	return renderer
+}
+
+// SetDefaultRenderer sets the default global renderer.
+func SetDefaultRenderer(r *Renderer) {
+	renderer = r
+}
+
+// NewRenderer creates a new Renderer.
+//
+// w will be used to determine the terminal's color capabilities.
+func NewRenderer(w io.Writer, opts ...termenv.OutputOption) *Renderer {
+	r := &Renderer{
+		output: termenv.NewOutput(w, opts...),
+	}
+	return r
+}
+
+// Output returns the termenv output.
+func (r *Renderer) Output() *termenv.Output {
+	r.mtx.RLock()
+	defer r.mtx.RUnlock()
+	return r.output
+}
+
+// SetOutput sets the termenv output.
+func (r *Renderer) SetOutput(o *termenv.Output) {
+	r.mtx.Lock()
+	defer r.mtx.Unlock()
+	r.output = o
+}
+
+// ColorProfile returns the detected termenv color profile.
+func (r *Renderer) ColorProfile() termenv.Profile {
+	r.mtx.RLock()
+	defer r.mtx.RUnlock()
+
+	if !r.explicitColorProfile {
+		r.getColorProfile.Do(func() {
+			// NOTE: we don't need to lock here because sync.Once provides its
+			// own locking mechanism.
+			r.colorProfile = r.output.EnvColorProfile()
+		})
+	}
+
+	return r.colorProfile
+}
+
+// ColorProfile returns the detected termenv color profile.
+func ColorProfile() termenv.Profile {
+	return renderer.ColorProfile()
+}
+
+// SetColorProfile sets the color profile on the renderer. This function exists
+// mostly for testing purposes so that you can assure you're testing against
+// a specific profile.
+//
+// Outside of testing you likely won't want to use this function as the color
+// profile will detect and cache the terminal's color capabilities and choose
+// the best available profile.
+//
+// Available color profiles are:
+//
+//	termenv.Ascii     // no color, 1-bit
+//	termenv.ANSI      //16 colors, 4-bit
+//	termenv.ANSI256   // 256 colors, 8-bit
+//	termenv.TrueColor // 16,777,216 colors, 24-bit
+//
+// This function is thread-safe.
+func (r *Renderer) SetColorProfile(p termenv.Profile) {
+	r.mtx.Lock()
+	defer r.mtx.Unlock()
+
+	r.colorProfile = p
+	r.explicitColorProfile = true
+}
+
+// SetColorProfile sets the color profile on the default renderer. This
+// function exists mostly for testing purposes so that you can assure you're
+// testing against a specific profile.
+//
+// Outside of testing you likely won't want to use this function as the color
+// profile will detect and cache the terminal's color capabilities and choose
+// the best available profile.
+//
+// Available color profiles are:
+//
+//	termenv.Ascii     // no color, 1-bit
+//	termenv.ANSI      //16 colors, 4-bit
+//	termenv.ANSI256   // 256 colors, 8-bit
+//	termenv.TrueColor // 16,777,216 colors, 24-bit
+//
+// This function is thread-safe.
+func SetColorProfile(p termenv.Profile) {
+	renderer.SetColorProfile(p)
+}
+
+// HasDarkBackground returns whether or not the terminal has a dark background.
+func HasDarkBackground() bool {
+	return renderer.HasDarkBackground()
+}
+
+// HasDarkBackground returns whether or not the renderer will render to a dark
+// background. A dark background can either be auto-detected, or set explicitly
+// on the renderer.
+func (r *Renderer) HasDarkBackground() bool {
+	r.mtx.RLock()
+	defer r.mtx.RUnlock()
+
+	if !r.explicitBackgroundColor {
+		r.getBackgroundColor.Do(func() {
+			// NOTE: we don't need to lock here because sync.Once provides its
+			// own locking mechanism.
+			r.hasDarkBackground = r.output.HasDarkBackground()
+		})
+	}
+
+	return r.hasDarkBackground
+}
+
+// SetHasDarkBackground sets the background color detection value for the
+// default renderer. This function exists mostly for testing purposes so that
+// you can assure you're testing against a specific background color setting.
+//
+// Outside of testing you likely won't want to use this function as the
+// backgrounds value will be automatically detected and cached against the
+// terminal's current background color setting.
+//
+// This function is thread-safe.
+func SetHasDarkBackground(b bool) {
+	renderer.SetHasDarkBackground(b)
+}
+
+// SetHasDarkBackground sets the background color detection value on the
+// renderer. This function exists mostly for testing purposes so that you can
+// assure you're testing against a specific background color setting.
+//
+// Outside of testing you likely won't want to use this function as the
+// backgrounds value will be automatically detected and cached against the
+// terminal's current background color setting.
+//
+// This function is thread-safe.
+func (r *Renderer) SetHasDarkBackground(b bool) {
+	r.mtx.Lock()
+	defer r.mtx.Unlock()
+
+	r.hasDarkBackground = b
+	r.explicitBackgroundColor = true
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/runes.go b/vendor/github.com/charmbracelet/lipgloss/runes.go
new file mode 100644
index 0000000..7a49e32
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/runes.go
@@ -0,0 +1,43 @@
+package lipgloss
+
+import (
+	"strings"
+)
+
+// StyleRunes apply a given style to runes at the given indices in the string.
+// Note that you must provide styling options for both matched and unmatched
+// runes. Indices out of bounds will be ignored.
+func StyleRunes(str string, indices []int, matched, unmatched Style) string {
+	// Convert slice of indices to a map for easier lookups
+	m := make(map[int]struct{})
+	for _, i := range indices {
+		m[i] = struct{}{}
+	}
+
+	var (
+		out   strings.Builder
+		group strings.Builder
+		style Style
+		runes = []rune(str)
+	)
+
+	for i, r := range runes {
+		group.WriteRune(r)
+
+		_, matches := m[i]
+		_, nextMatches := m[i+1]
+
+		if matches != nextMatches || i == len(runes)-1 {
+			// Flush
+			if matches {
+				style = matched
+			} else {
+				style = unmatched
+			}
+			out.WriteString(style.Render(group.String()))
+			group.Reset()
+		}
+	}
+
+	return out.String()
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/set.go b/vendor/github.com/charmbracelet/lipgloss/set.go
new file mode 100644
index 0000000..fde38fa
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/set.go
@@ -0,0 +1,799 @@
+package lipgloss
+
+// Set a value on the underlying rules map.
+func (s *Style) set(key propKey, value interface{}) {
+	// We don't allow negative integers on any of our other values, so just keep
+	// them at zero or above. We could use uints instead, but the
+	// conversions are a little tedious, so we're sticking with ints for
+	// sake of usability.
+	switch key { //nolint:exhaustive
+	case foregroundKey:
+		s.fgColor = colorOrNil(value)
+	case backgroundKey:
+		s.bgColor = colorOrNil(value)
+	case widthKey:
+		s.width = max(0, value.(int))
+	case heightKey:
+		s.height = max(0, value.(int))
+	case alignHorizontalKey:
+		s.alignHorizontal = value.(Position)
+	case alignVerticalKey:
+		s.alignVertical = value.(Position)
+	case paddingTopKey:
+		s.paddingTop = max(0, value.(int))
+	case paddingRightKey:
+		s.paddingRight = max(0, value.(int))
+	case paddingBottomKey:
+		s.paddingBottom = max(0, value.(int))
+	case paddingLeftKey:
+		s.paddingLeft = max(0, value.(int))
+	case marginTopKey:
+		s.marginTop = max(0, value.(int))
+	case marginRightKey:
+		s.marginRight = max(0, value.(int))
+	case marginBottomKey:
+		s.marginBottom = max(0, value.(int))
+	case marginLeftKey:
+		s.marginLeft = max(0, value.(int))
+	case marginBackgroundKey:
+		s.marginBgColor = colorOrNil(value)
+	case borderStyleKey:
+		s.borderStyle = value.(Border)
+	case borderTopForegroundKey:
+		s.borderTopFgColor = colorOrNil(value)
+	case borderRightForegroundKey:
+		s.borderRightFgColor = colorOrNil(value)
+	case borderBottomForegroundKey:
+		s.borderBottomFgColor = colorOrNil(value)
+	case borderLeftForegroundKey:
+		s.borderLeftFgColor = colorOrNil(value)
+	case borderTopBackgroundKey:
+		s.borderTopBgColor = colorOrNil(value)
+	case borderRightBackgroundKey:
+		s.borderRightBgColor = colorOrNil(value)
+	case borderBottomBackgroundKey:
+		s.borderBottomBgColor = colorOrNil(value)
+	case borderLeftBackgroundKey:
+		s.borderLeftBgColor = colorOrNil(value)
+	case maxWidthKey:
+		s.maxWidth = max(0, value.(int))
+	case maxHeightKey:
+		s.maxHeight = max(0, value.(int))
+	case tabWidthKey:
+		// TabWidth is the only property that may have a negative value (and
+		// that negative value can be no less than -1).
+		s.tabWidth = value.(int)
+	case transformKey:
+		s.transform = value.(func(string) string)
+	default:
+		if v, ok := value.(bool); ok { //nolint:nestif
+			if v {
+				s.attrs |= int(key)
+			} else {
+				s.attrs &^= int(key)
+			}
+		} else if attrs, ok := value.(int); ok {
+			// bool attrs
+			if attrs&int(key) != 0 {
+				s.attrs |= int(key)
+			} else {
+				s.attrs &^= int(key)
+			}
+		}
+	}
+
+	// Set the prop on
+	s.props = s.props.set(key)
+}
+
+// setFrom sets the property from another style.
+func (s *Style) setFrom(key propKey, i Style) {
+	switch key { //nolint:exhaustive
+	case foregroundKey:
+		s.set(foregroundKey, i.fgColor)
+	case backgroundKey:
+		s.set(backgroundKey, i.bgColor)
+	case widthKey:
+		s.set(widthKey, i.width)
+	case heightKey:
+		s.set(heightKey, i.height)
+	case alignHorizontalKey:
+		s.set(alignHorizontalKey, i.alignHorizontal)
+	case alignVerticalKey:
+		s.set(alignVerticalKey, i.alignVertical)
+	case paddingTopKey:
+		s.set(paddingTopKey, i.paddingTop)
+	case paddingRightKey:
+		s.set(paddingRightKey, i.paddingRight)
+	case paddingBottomKey:
+		s.set(paddingBottomKey, i.paddingBottom)
+	case paddingLeftKey:
+		s.set(paddingLeftKey, i.paddingLeft)
+	case marginTopKey:
+		s.set(marginTopKey, i.marginTop)
+	case marginRightKey:
+		s.set(marginRightKey, i.marginRight)
+	case marginBottomKey:
+		s.set(marginBottomKey, i.marginBottom)
+	case marginLeftKey:
+		s.set(marginLeftKey, i.marginLeft)
+	case marginBackgroundKey:
+		s.set(marginBackgroundKey, i.marginBgColor)
+	case borderStyleKey:
+		s.set(borderStyleKey, i.borderStyle)
+	case borderTopForegroundKey:
+		s.set(borderTopForegroundKey, i.borderTopFgColor)
+	case borderRightForegroundKey:
+		s.set(borderRightForegroundKey, i.borderRightFgColor)
+	case borderBottomForegroundKey:
+		s.set(borderBottomForegroundKey, i.borderBottomFgColor)
+	case borderLeftForegroundKey:
+		s.set(borderLeftForegroundKey, i.borderLeftFgColor)
+	case borderTopBackgroundKey:
+		s.set(borderTopBackgroundKey, i.borderTopBgColor)
+	case borderRightBackgroundKey:
+		s.set(borderRightBackgroundKey, i.borderRightBgColor)
+	case borderBottomBackgroundKey:
+		s.set(borderBottomBackgroundKey, i.borderBottomBgColor)
+	case borderLeftBackgroundKey:
+		s.set(borderLeftBackgroundKey, i.borderLeftBgColor)
+	case maxWidthKey:
+		s.set(maxWidthKey, i.maxWidth)
+	case maxHeightKey:
+		s.set(maxHeightKey, i.maxHeight)
+	case tabWidthKey:
+		s.set(tabWidthKey, i.tabWidth)
+	case transformKey:
+		s.set(transformKey, i.transform)
+	default:
+		// Set attributes for set bool properties
+		s.set(key, i.attrs)
+	}
+}
+
+func colorOrNil(c interface{}) TerminalColor {
+	if c, ok := c.(TerminalColor); ok {
+		return c
+	}
+	return nil
+}
+
+// Bold sets a bold formatting rule.
+func (s Style) Bold(v bool) Style {
+	s.set(boldKey, v)
+	return s
+}
+
+// Italic sets an italic formatting rule. In some terminal emulators this will
+// render with "reverse" coloring if not italic font variant is available.
+func (s Style) Italic(v bool) Style {
+	s.set(italicKey, v)
+	return s
+}
+
+// Underline sets an underline rule. By default, underlines will not be drawn on
+// whitespace like margins and padding. To change this behavior set
+// UnderlineSpaces.
+func (s Style) Underline(v bool) Style {
+	s.set(underlineKey, v)
+	return s
+}
+
+// Strikethrough sets a strikethrough rule. By default, strikes will not be
+// drawn on whitespace like margins and padding. To change this behavior set
+// StrikethroughSpaces.
+func (s Style) Strikethrough(v bool) Style {
+	s.set(strikethroughKey, v)
+	return s
+}
+
+// Reverse sets a rule for inverting foreground and background colors.
+func (s Style) Reverse(v bool) Style {
+	s.set(reverseKey, v)
+	return s
+}
+
+// Blink sets a rule for blinking foreground text.
+func (s Style) Blink(v bool) Style {
+	s.set(blinkKey, v)
+	return s
+}
+
+// Faint sets a rule for rendering the foreground color in a dimmer shade.
+func (s Style) Faint(v bool) Style {
+	s.set(faintKey, v)
+	return s
+}
+
+// Foreground sets a foreground color.
+//
+//	// Sets the foreground to blue
+//	s := lipgloss.NewStyle().Foreground(lipgloss.Color("#0000ff"))
+//
+//	// Removes the foreground color
+//	s.Foreground(lipgloss.NoColor)
+func (s Style) Foreground(c TerminalColor) Style {
+	s.set(foregroundKey, c)
+	return s
+}
+
+// Background sets a background color.
+func (s Style) Background(c TerminalColor) Style {
+	s.set(backgroundKey, c)
+	return s
+}
+
+// Width sets the width of the block before applying margins. The width, if
+// set, also determines where text will wrap.
+func (s Style) Width(i int) Style {
+	s.set(widthKey, i)
+	return s
+}
+
+// Height sets the height of the block before applying margins. If the height of
+// the text block is less than this value after applying padding (or not), the
+// block will be set to this height.
+func (s Style) Height(i int) Style {
+	s.set(heightKey, i)
+	return s
+}
+
+// Align is a shorthand method for setting horizontal and vertical alignment.
+//
+// With one argument, the position value is applied to the horizontal alignment.
+//
+// With two arguments, the value is applied to the horizontal and vertical
+// alignments, in that order.
+func (s Style) Align(p ...Position) Style {
+	if len(p) > 0 {
+		s.set(alignHorizontalKey, p[0])
+	}
+	if len(p) > 1 {
+		s.set(alignVerticalKey, p[1])
+	}
+	return s
+}
+
+// AlignHorizontal sets a horizontal text alignment rule.
+func (s Style) AlignHorizontal(p Position) Style {
+	s.set(alignHorizontalKey, p)
+	return s
+}
+
+// AlignVertical sets a vertical text alignment rule.
+func (s Style) AlignVertical(p Position) Style {
+	s.set(alignVerticalKey, p)
+	return s
+}
+
+// Padding is a shorthand method for setting padding on all sides at once.
+//
+// With one argument, the value is applied to all sides.
+//
+// With two arguments, the value is applied to the vertical and horizontal
+// sides, in that order.
+//
+// With three arguments, the value is applied to the top side, the horizontal
+// sides, and the bottom side, in that order.
+//
+// With four arguments, the value is applied clockwise starting from the top
+// side, followed by the right side, then the bottom, and finally the left.
+//
+// With more than four arguments no padding will be added.
+func (s Style) Padding(i ...int) Style {
+	top, right, bottom, left, ok := whichSidesInt(i...)
+	if !ok {
+		return s
+	}
+
+	s.set(paddingTopKey, top)
+	s.set(paddingRightKey, right)
+	s.set(paddingBottomKey, bottom)
+	s.set(paddingLeftKey, left)
+	return s
+}
+
+// PaddingLeft adds padding on the left.
+func (s Style) PaddingLeft(i int) Style {
+	s.set(paddingLeftKey, i)
+	return s
+}
+
+// PaddingRight adds padding on the right.
+func (s Style) PaddingRight(i int) Style {
+	s.set(paddingRightKey, i)
+	return s
+}
+
+// PaddingTop adds padding to the top of the block.
+func (s Style) PaddingTop(i int) Style {
+	s.set(paddingTopKey, i)
+	return s
+}
+
+// PaddingBottom adds padding to the bottom of the block.
+func (s Style) PaddingBottom(i int) Style {
+	s.set(paddingBottomKey, i)
+	return s
+}
+
+// ColorWhitespace determines whether or not the background color should be
+// applied to the padding. This is true by default as it's more than likely the
+// desired and expected behavior, but it can be disabled for certain graphic
+// effects.
+//
+// Deprecated: Just use margins and padding.
+func (s Style) ColorWhitespace(v bool) Style {
+	s.set(colorWhitespaceKey, v)
+	return s
+}
+
+// Margin is a shorthand method for setting margins on all sides at once.
+//
+// With one argument, the value is applied to all sides.
+//
+// With two arguments, the value is applied to the vertical and horizontal
+// sides, in that order.
+//
+// With three arguments, the value is applied to the top side, the horizontal
+// sides, and the bottom side, in that order.
+//
+// With four arguments, the value is applied clockwise starting from the top
+// side, followed by the right side, then the bottom, and finally the left.
+//
+// With more than four arguments no margin will be added.
+func (s Style) Margin(i ...int) Style {
+	top, right, bottom, left, ok := whichSidesInt(i...)
+	if !ok {
+		return s
+	}
+
+	s.set(marginTopKey, top)
+	s.set(marginRightKey, right)
+	s.set(marginBottomKey, bottom)
+	s.set(marginLeftKey, left)
+	return s
+}
+
+// MarginLeft sets the value of the left margin.
+func (s Style) MarginLeft(i int) Style {
+	s.set(marginLeftKey, i)
+	return s
+}
+
+// MarginRight sets the value of the right margin.
+func (s Style) MarginRight(i int) Style {
+	s.set(marginRightKey, i)
+	return s
+}
+
+// MarginTop sets the value of the top margin.
+func (s Style) MarginTop(i int) Style {
+	s.set(marginTopKey, i)
+	return s
+}
+
+// MarginBottom sets the value of the bottom margin.
+func (s Style) MarginBottom(i int) Style {
+	s.set(marginBottomKey, i)
+	return s
+}
+
+// MarginBackground sets the background color of the margin. Note that this is
+// also set when inheriting from a style with a background color. In that case
+// the background color on that style will set the margin color on this style.
+func (s Style) MarginBackground(c TerminalColor) Style {
+	s.set(marginBackgroundKey, c)
+	return s
+}
+
+// Border is shorthand for setting the border style and which sides should
+// have a border at once. The variadic argument sides works as follows:
+//
+// With one value, the value is applied to all sides.
+//
+// With two values, the values are applied to the vertical and horizontal
+// sides, in that order.
+//
+// With three values, the values are applied to the top side, the horizontal
+// sides, and the bottom side, in that order.
+//
+// With four values, the values are applied clockwise starting from the top
+// side, followed by the right side, then the bottom, and finally the left.
+//
+// With more than four arguments the border will be applied to all sides.
+//
+// Examples:
+//
+//	// Applies borders to the top and bottom only
+//	lipgloss.NewStyle().Border(lipgloss.NormalBorder(), true, false)
+//
+//	// Applies rounded borders to the right and bottom only
+//	lipgloss.NewStyle().Border(lipgloss.RoundedBorder(), false, true, true, false)
+func (s Style) Border(b Border, sides ...bool) Style {
+	s.set(borderStyleKey, b)
+
+	top, right, bottom, left, ok := whichSidesBool(sides...)
+	if !ok {
+		top = true
+		right = true
+		bottom = true
+		left = true
+	}
+
+	s.set(borderTopKey, top)
+	s.set(borderRightKey, right)
+	s.set(borderBottomKey, bottom)
+	s.set(borderLeftKey, left)
+
+	return s
+}
+
+// BorderStyle defines the Border on a style. A Border contains a series of
+// definitions for the sides and corners of a border.
+//
+// Note that if border visibility has not been set for any sides when setting
+// the border style, the border will be enabled for all sides during rendering.
+//
+// You can define border characters as you'd like, though several default
+// styles are included: NormalBorder(), RoundedBorder(), BlockBorder(),
+// OuterHalfBlockBorder(), InnerHalfBlockBorder(), ThickBorder(),
+// and DoubleBorder().
+//
+// Example:
+//
+//	lipgloss.NewStyle().BorderStyle(lipgloss.ThickBorder())
+func (s Style) BorderStyle(b Border) Style {
+	s.set(borderStyleKey, b)
+	return s
+}
+
+// BorderTop determines whether or not to draw a top border.
+func (s Style) BorderTop(v bool) Style {
+	s.set(borderTopKey, v)
+	return s
+}
+
+// BorderRight determines whether or not to draw a right border.
+func (s Style) BorderRight(v bool) Style {
+	s.set(borderRightKey, v)
+	return s
+}
+
+// BorderBottom determines whether or not to draw a bottom border.
+func (s Style) BorderBottom(v bool) Style {
+	s.set(borderBottomKey, v)
+	return s
+}
+
+// BorderLeft determines whether or not to draw a left border.
+func (s Style) BorderLeft(v bool) Style {
+	s.set(borderLeftKey, v)
+	return s
+}
+
+// BorderForeground is a shorthand function for setting all of the
+// foreground colors of the borders at once. The arguments work as follows:
+//
+// With one argument, the argument is applied to all sides.
+//
+// With two arguments, the arguments are applied to the vertical and horizontal
+// sides, in that order.
+//
+// With three arguments, the arguments are applied to the top side, the
+// horizontal sides, and the bottom side, in that order.
+//
+// With four arguments, the arguments are applied clockwise starting from the
+// top side, followed by the right side, then the bottom, and finally the left.
+//
+// With more than four arguments nothing will be set.
+func (s Style) BorderForeground(c ...TerminalColor) Style {
+	if len(c) == 0 {
+		return s
+	}
+
+	top, right, bottom, left, ok := whichSidesColor(c...)
+	if !ok {
+		return s
+	}
+
+	s.set(borderTopForegroundKey, top)
+	s.set(borderRightForegroundKey, right)
+	s.set(borderBottomForegroundKey, bottom)
+	s.set(borderLeftForegroundKey, left)
+
+	return s
+}
+
+// BorderTopForeground set the foreground color for the top of the border.
+func (s Style) BorderTopForeground(c TerminalColor) Style {
+	s.set(borderTopForegroundKey, c)
+	return s
+}
+
+// BorderRightForeground sets the foreground color for the right side of the
+// border.
+func (s Style) BorderRightForeground(c TerminalColor) Style {
+	s.set(borderRightForegroundKey, c)
+	return s
+}
+
+// BorderBottomForeground sets the foreground color for the bottom of the
+// border.
+func (s Style) BorderBottomForeground(c TerminalColor) Style {
+	s.set(borderBottomForegroundKey, c)
+	return s
+}
+
+// BorderLeftForeground sets the foreground color for the left side of the
+// border.
+func (s Style) BorderLeftForeground(c TerminalColor) Style {
+	s.set(borderLeftForegroundKey, c)
+	return s
+}
+
+// BorderBackground is a shorthand function for setting all of the
+// background colors of the borders at once. The arguments work as follows:
+//
+// With one argument, the argument is applied to all sides.
+//
+// With two arguments, the arguments are applied to the vertical and horizontal
+// sides, in that order.
+//
+// With three arguments, the arguments are applied to the top side, the
+// horizontal sides, and the bottom side, in that order.
+//
+// With four arguments, the arguments are applied clockwise starting from the
+// top side, followed by the right side, then the bottom, and finally the left.
+//
+// With more than four arguments nothing will be set.
+func (s Style) BorderBackground(c ...TerminalColor) Style {
+	if len(c) == 0 {
+		return s
+	}
+
+	top, right, bottom, left, ok := whichSidesColor(c...)
+	if !ok {
+		return s
+	}
+
+	s.set(borderTopBackgroundKey, top)
+	s.set(borderRightBackgroundKey, right)
+	s.set(borderBottomBackgroundKey, bottom)
+	s.set(borderLeftBackgroundKey, left)
+
+	return s
+}
+
+// BorderTopBackground sets the background color of the top of the border.
+func (s Style) BorderTopBackground(c TerminalColor) Style {
+	s.set(borderTopBackgroundKey, c)
+	return s
+}
+
+// BorderRightBackground sets the background color of right side the border.
+func (s Style) BorderRightBackground(c TerminalColor) Style {
+	s.set(borderRightBackgroundKey, c)
+	return s
+}
+
+// BorderBottomBackground sets the background color of the bottom of the
+// border.
+func (s Style) BorderBottomBackground(c TerminalColor) Style {
+	s.set(borderBottomBackgroundKey, c)
+	return s
+}
+
+// BorderLeftBackground set the background color of the left side of the
+// border.
+func (s Style) BorderLeftBackground(c TerminalColor) Style {
+	s.set(borderLeftBackgroundKey, c)
+	return s
+}
+
+// Inline makes rendering output one line and disables the rendering of
+// margins, padding and borders. This is useful when you need a style to apply
+// only to font rendering and don't want it to change any physical dimensions.
+// It works well with Style.MaxWidth.
+//
+// Because this in intended to be used at the time of render, this method will
+// not mutate the style and instead return a copy.
+//
+// Example:
+//
+//	var userInput string = "..."
+//	var userStyle = text.Style{ /* ... */ }
+//	fmt.Println(userStyle.Inline(true).Render(userInput))
+func (s Style) Inline(v bool) Style {
+	o := s // copy
+	o.set(inlineKey, v)
+	return o
+}
+
+// MaxWidth applies a max width to a given style. This is useful in enforcing
+// a certain width at render time, particularly with arbitrary strings and
+// styles.
+//
+// Because this in intended to be used at the time of render, this method will
+// not mutate the style and instead return a copy.
+//
+// Example:
+//
+//	var userInput string = "..."
+//	var userStyle = text.Style{ /* ... */ }
+//	fmt.Println(userStyle.MaxWidth(16).Render(userInput))
+func (s Style) MaxWidth(n int) Style {
+	o := s // copy
+	o.set(maxWidthKey, n)
+	return o
+}
+
+// MaxHeight applies a max height to a given style. This is useful in enforcing
+// a certain height at render time, particularly with arbitrary strings and
+// styles.
+//
+// Because this in intended to be used at the time of render, this method will
+// not mutate the style and instead returns a copy.
+func (s Style) MaxHeight(n int) Style {
+	o := s // copy
+	o.set(maxHeightKey, n)
+	return o
+}
+
+// NoTabConversion can be passed to [Style.TabWidth] to disable the replacement
+// of tabs with spaces at render time.
+const NoTabConversion = -1
+
+// TabWidth sets the number of spaces that a tab (/t) should be rendered as.
+// When set to 0, tabs will be removed. To disable the replacement of tabs with
+// spaces entirely, set this to [NoTabConversion].
+//
+// By default, tabs will be replaced with 4 spaces.
+func (s Style) TabWidth(n int) Style {
+	if n <= -1 {
+		n = -1
+	}
+	s.set(tabWidthKey, n)
+	return s
+}
+
+// UnderlineSpaces determines whether to underline spaces between words. By
+// default, this is true. Spaces can also be underlined without underlining the
+// text itself.
+func (s Style) UnderlineSpaces(v bool) Style {
+	s.set(underlineSpacesKey, v)
+	return s
+}
+
+// StrikethroughSpaces determines whether to apply strikethroughs to spaces
+// between words. By default, this is true. Spaces can also be struck without
+// underlining the text itself.
+func (s Style) StrikethroughSpaces(v bool) Style {
+	s.set(strikethroughSpacesKey, v)
+	return s
+}
+
+// Transform applies a given function to a string at render time, allowing for
+// the string being rendered to be manipuated.
+//
+// Example:
+//
+//	s := NewStyle().Transform(strings.ToUpper)
+//	fmt.Println(s.Render("raow!") // "RAOW!"
+func (s Style) Transform(fn func(string) string) Style {
+	s.set(transformKey, fn)
+	return s
+}
+
+// Renderer sets the renderer for the style. This is useful for changing the
+// renderer for a style that is being used in a different context.
+func (s Style) Renderer(r *Renderer) Style {
+	s.r = r
+	return s
+}
+
+// whichSidesInt is a helper method for setting values on sides of a block based
+// on the number of arguments. It follows the CSS shorthand rules for blocks
+// like margin, padding. and borders. Here are how the rules work:
+//
+// 0 args:  do nothing
+// 1 arg:   all sides
+// 2 args:  top -> bottom
+// 3 args:  top -> horizontal -> bottom
+// 4 args:  top -> right -> bottom -> left
+// 5+ args: do nothing.
+func whichSidesInt(i ...int) (top, right, bottom, left int, ok bool) {
+	switch len(i) {
+	case 1:
+		top = i[0]
+		bottom = i[0]
+		left = i[0]
+		right = i[0]
+		ok = true
+	case 2: //nolint:mnd
+		top = i[0]
+		bottom = i[0]
+		left = i[1]
+		right = i[1]
+		ok = true
+	case 3: //nolint:mnd
+		top = i[0]
+		left = i[1]
+		right = i[1]
+		bottom = i[2]
+		ok = true
+	case 4: //nolint:mnd
+		top = i[0]
+		right = i[1]
+		bottom = i[2]
+		left = i[3]
+		ok = true
+	}
+	return top, right, bottom, left, ok
+}
+
+// whichSidesBool is like whichSidesInt, except it operates on a series of
+// boolean values. See the comment on whichSidesInt for details on how this
+// works.
+func whichSidesBool(i ...bool) (top, right, bottom, left bool, ok bool) {
+	switch len(i) {
+	case 1:
+		top = i[0]
+		bottom = i[0]
+		left = i[0]
+		right = i[0]
+		ok = true
+	case 2: //nolint:mnd
+		top = i[0]
+		bottom = i[0]
+		left = i[1]
+		right = i[1]
+		ok = true
+	case 3: //nolint:mnd
+		top = i[0]
+		left = i[1]
+		right = i[1]
+		bottom = i[2]
+		ok = true
+	case 4: //nolint:mnd
+		top = i[0]
+		right = i[1]
+		bottom = i[2]
+		left = i[3]
+		ok = true
+	}
+	return top, right, bottom, left, ok
+}
+
+// whichSidesColor is like whichSides, except it operates on a series of
+// boolean values. See the comment on whichSidesInt for details on how this
+// works.
+func whichSidesColor(i ...TerminalColor) (top, right, bottom, left TerminalColor, ok bool) {
+	switch len(i) {
+	case 1:
+		top = i[0]
+		bottom = i[0]
+		left = i[0]
+		right = i[0]
+		ok = true
+	case 2: //nolint:mnd
+		top = i[0]
+		bottom = i[0]
+		left = i[1]
+		right = i[1]
+		ok = true
+	case 3: //nolint:mnd
+		top = i[0]
+		left = i[1]
+		right = i[1]
+		bottom = i[2]
+		ok = true
+	case 4: //nolint:mnd
+		top = i[0]
+		right = i[1]
+		bottom = i[2]
+		left = i[3]
+		ok = true
+	}
+	return top, right, bottom, left, ok
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/size.go b/vendor/github.com/charmbracelet/lipgloss/size.go
new file mode 100644
index 0000000..e169ff5
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/size.go
@@ -0,0 +1,41 @@
+package lipgloss
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// Width returns the cell width of characters in the string. ANSI sequences are
+// ignored and characters wider than one cell (such as Chinese characters and
+// emojis) are appropriately measured.
+//
+// You should use this instead of len(string) len([]rune(string) as neither
+// will give you accurate results.
+func Width(str string) (width int) {
+	for _, l := range strings.Split(str, "\n") {
+		w := ansi.StringWidth(l)
+		if w > width {
+			width = w
+		}
+	}
+
+	return width
+}
+
+// Height returns height of a string in cells. This is done simply by
+// counting \n characters. If your strings use \r\n for newlines you should
+// convert them to \n first, or simply write a separate function for measuring
+// height.
+func Height(str string) int {
+	return strings.Count(str, "\n") + 1
+}
+
+// Size returns the width and height of the string in cells. ANSI sequences are
+// ignored and characters wider than one cell (such as Chinese characters and
+// emojis) are appropriately measured.
+func Size(str string) (width, height int) {
+	width = Width(str)
+	height = Height(str)
+	return width, height
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/style.go b/vendor/github.com/charmbracelet/lipgloss/style.go
new file mode 100644
index 0000000..59fa3ab
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/style.go
@@ -0,0 +1,588 @@
+package lipgloss
+
+import (
+	"strings"
+	"unicode"
+
+	"github.com/charmbracelet/x/ansi"
+	"github.com/charmbracelet/x/cellbuf"
+	"github.com/muesli/termenv"
+)
+
+const tabWidthDefault = 4
+
+// Property for a key.
+type propKey int64
+
+// Available properties.
+const (
+	// Boolean props come first.
+	boldKey propKey = 1 << iota
+	italicKey
+	underlineKey
+	strikethroughKey
+	reverseKey
+	blinkKey
+	faintKey
+	underlineSpacesKey
+	strikethroughSpacesKey
+	colorWhitespaceKey
+
+	// Non-boolean props.
+	foregroundKey
+	backgroundKey
+	widthKey
+	heightKey
+	alignHorizontalKey
+	alignVerticalKey
+
+	// Padding.
+	paddingTopKey
+	paddingRightKey
+	paddingBottomKey
+	paddingLeftKey
+
+	// Margins.
+	marginTopKey
+	marginRightKey
+	marginBottomKey
+	marginLeftKey
+	marginBackgroundKey
+
+	// Border runes.
+	borderStyleKey
+
+	// Border edges.
+	borderTopKey
+	borderRightKey
+	borderBottomKey
+	borderLeftKey
+
+	// Border foreground colors.
+	borderTopForegroundKey
+	borderRightForegroundKey
+	borderBottomForegroundKey
+	borderLeftForegroundKey
+
+	// Border background colors.
+	borderTopBackgroundKey
+	borderRightBackgroundKey
+	borderBottomBackgroundKey
+	borderLeftBackgroundKey
+
+	inlineKey
+	maxWidthKey
+	maxHeightKey
+	tabWidthKey
+
+	transformKey
+)
+
+// props is a set of properties.
+type props int64
+
+// set sets a property.
+func (p props) set(k propKey) props {
+	return p | props(k)
+}
+
+// unset unsets a property.
+func (p props) unset(k propKey) props {
+	return p &^ props(k)
+}
+
+// has checks if a property is set.
+func (p props) has(k propKey) bool {
+	return p&props(k) != 0
+}
+
+// NewStyle returns a new, empty Style. While it's syntactic sugar for the
+// Style{} primitive, it's recommended to use this function for creating styles
+// in case the underlying implementation changes. It takes an optional string
+// value to be set as the underlying string value for this style.
+func NewStyle() Style {
+	return renderer.NewStyle()
+}
+
+// NewStyle returns a new, empty Style. While it's syntactic sugar for the
+// Style{} primitive, it's recommended to use this function for creating styles
+// in case the underlying implementation changes. It takes an optional string
+// value to be set as the underlying string value for this style.
+func (r *Renderer) NewStyle() Style {
+	s := Style{r: r}
+	return s
+}
+
+// Style contains a set of rules that comprise a style as a whole.
+type Style struct {
+	r     *Renderer
+	props props
+	value string
+
+	// we store bool props values here
+	attrs int
+
+	// props that have values
+	fgColor TerminalColor
+	bgColor TerminalColor
+
+	width  int
+	height int
+
+	alignHorizontal Position
+	alignVertical   Position
+
+	paddingTop    int
+	paddingRight  int
+	paddingBottom int
+	paddingLeft   int
+
+	marginTop     int
+	marginRight   int
+	marginBottom  int
+	marginLeft    int
+	marginBgColor TerminalColor
+
+	borderStyle         Border
+	borderTopFgColor    TerminalColor
+	borderRightFgColor  TerminalColor
+	borderBottomFgColor TerminalColor
+	borderLeftFgColor   TerminalColor
+	borderTopBgColor    TerminalColor
+	borderRightBgColor  TerminalColor
+	borderBottomBgColor TerminalColor
+	borderLeftBgColor   TerminalColor
+
+	maxWidth  int
+	maxHeight int
+	tabWidth  int
+
+	transform func(string) string
+}
+
+// joinString joins a list of strings into a single string separated with a
+// space.
+func joinString(strs ...string) string {
+	return strings.Join(strs, " ")
+}
+
+// SetString sets the underlying string value for this style. To render once
+// the underlying string is set, use the Style.String. This method is
+// a convenience for cases when having a stringer implementation is handy, such
+// as when using fmt.Sprintf. You can also simply define a style and render out
+// strings directly with Style.Render.
+func (s Style) SetString(strs ...string) Style {
+	s.value = joinString(strs...)
+	return s
+}
+
+// Value returns the raw, unformatted, underlying string value for this style.
+func (s Style) Value() string {
+	return s.value
+}
+
+// String implements stringer for a Style, returning the rendered result based
+// on the rules in this style. An underlying string value must be set with
+// Style.SetString prior to using this method.
+func (s Style) String() string {
+	return s.Render()
+}
+
+// Copy returns a copy of this style, including any underlying string values.
+//
+// Deprecated: to copy just use assignment (i.e. a := b). All methods also
+// return a new style.
+func (s Style) Copy() Style {
+	return s
+}
+
+// Inherit overlays the style in the argument onto this style by copying each explicitly
+// set value from the argument style onto this style if it is not already explicitly set.
+// Existing set values are kept intact and not overwritten.
+//
+// Margins, padding, and underlying string values are not inherited.
+func (s Style) Inherit(i Style) Style {
+	for k := boldKey; k <= transformKey; k <<= 1 {
+		if !i.isSet(k) {
+			continue
+		}
+
+		switch k { //nolint:exhaustive
+		case marginTopKey, marginRightKey, marginBottomKey, marginLeftKey:
+			// Margins are not inherited
+			continue
+		case paddingTopKey, paddingRightKey, paddingBottomKey, paddingLeftKey:
+			// Padding is not inherited
+			continue
+		case backgroundKey:
+			// The margins also inherit the background color
+			if !s.isSet(marginBackgroundKey) && !i.isSet(marginBackgroundKey) {
+				s.set(marginBackgroundKey, i.bgColor)
+			}
+		}
+
+		if s.isSet(k) {
+			continue
+		}
+
+		s.setFrom(k, i)
+	}
+	return s
+}
+
+// Render applies the defined style formatting to a given string.
+func (s Style) Render(strs ...string) string {
+	if s.r == nil {
+		s.r = renderer
+	}
+	if s.value != "" {
+		strs = append([]string{s.value}, strs...)
+	}
+
+	var (
+		str = joinString(strs...)
+
+		p            = s.r.ColorProfile()
+		te           = p.String()
+		teSpace      = p.String()
+		teWhitespace = p.String()
+
+		bold          = s.getAsBool(boldKey, false)
+		italic        = s.getAsBool(italicKey, false)
+		underline     = s.getAsBool(underlineKey, false)
+		strikethrough = s.getAsBool(strikethroughKey, false)
+		reverse       = s.getAsBool(reverseKey, false)
+		blink         = s.getAsBool(blinkKey, false)
+		faint         = s.getAsBool(faintKey, false)
+
+		fg = s.getAsColor(foregroundKey)
+		bg = s.getAsColor(backgroundKey)
+
+		width           = s.getAsInt(widthKey)
+		height          = s.getAsInt(heightKey)
+		horizontalAlign = s.getAsPosition(alignHorizontalKey)
+		verticalAlign   = s.getAsPosition(alignVerticalKey)
+
+		topPadding    = s.getAsInt(paddingTopKey)
+		rightPadding  = s.getAsInt(paddingRightKey)
+		bottomPadding = s.getAsInt(paddingBottomKey)
+		leftPadding   = s.getAsInt(paddingLeftKey)
+
+		colorWhitespace = s.getAsBool(colorWhitespaceKey, true)
+		inline          = s.getAsBool(inlineKey, false)
+		maxWidth        = s.getAsInt(maxWidthKey)
+		maxHeight       = s.getAsInt(maxHeightKey)
+
+		underlineSpaces     = s.getAsBool(underlineSpacesKey, false) || (underline && s.getAsBool(underlineSpacesKey, true))
+		strikethroughSpaces = s.getAsBool(strikethroughSpacesKey, false) || (strikethrough && s.getAsBool(strikethroughSpacesKey, true))
+
+		// Do we need to style whitespace (padding and space outside
+		// paragraphs) separately?
+		styleWhitespace = reverse
+
+		// Do we need to style spaces separately?
+		useSpaceStyler = (underline && !underlineSpaces) || (strikethrough && !strikethroughSpaces) || underlineSpaces || strikethroughSpaces
+
+		transform = s.getAsTransform(transformKey)
+	)
+
+	if transform != nil {
+		str = transform(str)
+	}
+
+	if s.props == 0 {
+		return s.maybeConvertTabs(str)
+	}
+
+	// Enable support for ANSI on the legacy Windows cmd.exe console. This is a
+	// no-op on non-Windows systems and on Windows runs only once.
+	enableLegacyWindowsANSI()
+
+	if bold {
+		te = te.Bold()
+	}
+	if italic {
+		te = te.Italic()
+	}
+	if underline {
+		te = te.Underline()
+	}
+	if reverse {
+		teWhitespace = teWhitespace.Reverse()
+		te = te.Reverse()
+	}
+	if blink {
+		te = te.Blink()
+	}
+	if faint {
+		te = te.Faint()
+	}
+
+	if fg != noColor {
+		te = te.Foreground(fg.color(s.r))
+		if styleWhitespace {
+			teWhitespace = teWhitespace.Foreground(fg.color(s.r))
+		}
+		if useSpaceStyler {
+			teSpace = teSpace.Foreground(fg.color(s.r))
+		}
+	}
+
+	if bg != noColor {
+		te = te.Background(bg.color(s.r))
+		if colorWhitespace {
+			teWhitespace = teWhitespace.Background(bg.color(s.r))
+		}
+		if useSpaceStyler {
+			teSpace = teSpace.Background(bg.color(s.r))
+		}
+	}
+
+	if underline {
+		te = te.Underline()
+	}
+	if strikethrough {
+		te = te.CrossOut()
+	}
+
+	if underlineSpaces {
+		teSpace = teSpace.Underline()
+	}
+	if strikethroughSpaces {
+		teSpace = teSpace.CrossOut()
+	}
+
+	// Potentially convert tabs to spaces
+	str = s.maybeConvertTabs(str)
+	// carriage returns can cause strange behaviour when rendering.
+	str = strings.ReplaceAll(str, "\r\n", "\n")
+
+	// Strip newlines in single line mode
+	if inline {
+		str = strings.ReplaceAll(str, "\n", "")
+	}
+
+	// Word wrap
+	if !inline && width > 0 {
+		wrapAt := width - leftPadding - rightPadding
+		str = cellbuf.Wrap(str, wrapAt, "")
+	}
+
+	// Render core text
+	{
+		var b strings.Builder
+
+		l := strings.Split(str, "\n")
+		for i := range l {
+			if useSpaceStyler {
+				// Look for spaces and apply a different styler
+				for _, r := range l[i] {
+					if unicode.IsSpace(r) {
+						b.WriteString(teSpace.Styled(string(r)))
+						continue
+					}
+					b.WriteString(te.Styled(string(r)))
+				}
+			} else {
+				b.WriteString(te.Styled(l[i]))
+			}
+			if i != len(l)-1 {
+				b.WriteRune('\n')
+			}
+		}
+
+		str = b.String()
+	}
+
+	// Padding
+	if !inline { //nolint:nestif
+		if leftPadding > 0 {
+			var st *termenv.Style
+			if colorWhitespace || styleWhitespace {
+				st = &teWhitespace
+			}
+			str = padLeft(str, leftPadding, st)
+		}
+
+		if rightPadding > 0 {
+			var st *termenv.Style
+			if colorWhitespace || styleWhitespace {
+				st = &teWhitespace
+			}
+			str = padRight(str, rightPadding, st)
+		}
+
+		if topPadding > 0 {
+			str = strings.Repeat("\n", topPadding) + str
+		}
+
+		if bottomPadding > 0 {
+			str += strings.Repeat("\n", bottomPadding)
+		}
+	}
+
+	// Height
+	if height > 0 {
+		str = alignTextVertical(str, verticalAlign, height, nil)
+	}
+
+	// Set alignment. This will also pad short lines with spaces so that all
+	// lines are the same length, so we run it under a few different conditions
+	// beyond alignment.
+	{
+		numLines := strings.Count(str, "\n")
+
+		if numLines != 0 || width != 0 {
+			var st *termenv.Style
+			if colorWhitespace || styleWhitespace {
+				st = &teWhitespace
+			}
+			str = alignTextHorizontal(str, horizontalAlign, width, st)
+		}
+	}
+
+	if !inline {
+		str = s.applyBorder(str)
+		str = s.applyMargins(str, inline)
+	}
+
+	// Truncate according to MaxWidth
+	if maxWidth > 0 {
+		lines := strings.Split(str, "\n")
+
+		for i := range lines {
+			lines[i] = ansi.Truncate(lines[i], maxWidth, "")
+		}
+
+		str = strings.Join(lines, "\n")
+	}
+
+	// Truncate according to MaxHeight
+	if maxHeight > 0 {
+		lines := strings.Split(str, "\n")
+		height := min(maxHeight, len(lines))
+		if len(lines) > 0 {
+			str = strings.Join(lines[:height], "\n")
+		}
+	}
+
+	return str
+}
+
+func (s Style) maybeConvertTabs(str string) string {
+	tw := tabWidthDefault
+	if s.isSet(tabWidthKey) {
+		tw = s.getAsInt(tabWidthKey)
+	}
+	switch tw {
+	case -1:
+		return str
+	case 0:
+		return strings.ReplaceAll(str, "\t", "")
+	default:
+		return strings.ReplaceAll(str, "\t", strings.Repeat(" ", tw))
+	}
+}
+
+func (s Style) applyMargins(str string, inline bool) string {
+	var (
+		topMargin    = s.getAsInt(marginTopKey)
+		rightMargin  = s.getAsInt(marginRightKey)
+		bottomMargin = s.getAsInt(marginBottomKey)
+		leftMargin   = s.getAsInt(marginLeftKey)
+
+		styler termenv.Style
+	)
+
+	bgc := s.getAsColor(marginBackgroundKey)
+	if bgc != noColor {
+		styler = styler.Background(bgc.color(s.r))
+	}
+
+	// Add left and right margin
+	str = padLeft(str, leftMargin, &styler)
+	str = padRight(str, rightMargin, &styler)
+
+	// Top/bottom margin
+	if !inline {
+		_, width := getLines(str)
+		spaces := strings.Repeat(" ", width)
+
+		if topMargin > 0 {
+			str = styler.Styled(strings.Repeat(spaces+"\n", topMargin)) + str
+		}
+		if bottomMargin > 0 {
+			str += styler.Styled(strings.Repeat("\n"+spaces, bottomMargin))
+		}
+	}
+
+	return str
+}
+
+// Apply left padding.
+func padLeft(str string, n int, style *termenv.Style) string {
+	return pad(str, -n, style)
+}
+
+// Apply right padding.
+func padRight(str string, n int, style *termenv.Style) string {
+	return pad(str, n, style)
+}
+
+// pad adds padding to either the left or right side of a string.
+// Positive values add to the right side while negative values
+// add to the left side.
+func pad(str string, n int, style *termenv.Style) string {
+	if n == 0 {
+		return str
+	}
+
+	sp := strings.Repeat(" ", abs(n))
+	if style != nil {
+		sp = style.Styled(sp)
+	}
+
+	b := strings.Builder{}
+	l := strings.Split(str, "\n")
+
+	for i := range l {
+		switch {
+		// pad right
+		case n > 0:
+			b.WriteString(l[i])
+			b.WriteString(sp)
+		// pad left
+		default:
+			b.WriteString(sp)
+			b.WriteString(l[i])
+		}
+
+		if i != len(l)-1 {
+			b.WriteRune('\n')
+		}
+	}
+
+	return b.String()
+}
+
+func max(a, b int) int { //nolint:unparam,predeclared
+	if a > b {
+		return a
+	}
+	return b
+}
+
+func min(a, b int) int { //nolint:predeclared
+	if a < b {
+		return a
+	}
+	return b
+}
+
+func abs(a int) int {
+	if a < 0 {
+		return -a
+	}
+
+	return a
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/unset.go b/vendor/github.com/charmbracelet/lipgloss/unset.go
new file mode 100644
index 0000000..1086e72
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/unset.go
@@ -0,0 +1,331 @@
+package lipgloss
+
+// unset unsets a property from a style.
+func (s *Style) unset(key propKey) {
+	s.props = s.props.unset(key)
+}
+
+// UnsetBold removes the bold style rule, if set.
+func (s Style) UnsetBold() Style {
+	s.unset(boldKey)
+	return s
+}
+
+// UnsetItalic removes the italic style rule, if set.
+func (s Style) UnsetItalic() Style {
+	s.unset(italicKey)
+	return s
+}
+
+// UnsetUnderline removes the underline style rule, if set.
+func (s Style) UnsetUnderline() Style {
+	s.unset(underlineKey)
+	return s
+}
+
+// UnsetStrikethrough removes the strikethrough style rule, if set.
+func (s Style) UnsetStrikethrough() Style {
+	s.unset(strikethroughKey)
+	return s
+}
+
+// UnsetReverse removes the reverse style rule, if set.
+func (s Style) UnsetReverse() Style {
+	s.unset(reverseKey)
+	return s
+}
+
+// UnsetBlink removes the blink style rule, if set.
+func (s Style) UnsetBlink() Style {
+	s.unset(blinkKey)
+	return s
+}
+
+// UnsetFaint removes the faint style rule, if set.
+func (s Style) UnsetFaint() Style {
+	s.unset(faintKey)
+	return s
+}
+
+// UnsetForeground removes the foreground style rule, if set.
+func (s Style) UnsetForeground() Style {
+	s.unset(foregroundKey)
+	return s
+}
+
+// UnsetBackground removes the background style rule, if set.
+func (s Style) UnsetBackground() Style {
+	s.unset(backgroundKey)
+	return s
+}
+
+// UnsetWidth removes the width style rule, if set.
+func (s Style) UnsetWidth() Style {
+	s.unset(widthKey)
+	return s
+}
+
+// UnsetHeight removes the height style rule, if set.
+func (s Style) UnsetHeight() Style {
+	s.unset(heightKey)
+	return s
+}
+
+// UnsetAlign removes the horizontal and vertical text alignment style rule, if set.
+func (s Style) UnsetAlign() Style {
+	s.unset(alignHorizontalKey)
+	s.unset(alignVerticalKey)
+	return s
+}
+
+// UnsetAlignHorizontal removes the horizontal text alignment style rule, if set.
+func (s Style) UnsetAlignHorizontal() Style {
+	s.unset(alignHorizontalKey)
+	return s
+}
+
+// UnsetAlignVertical removes the vertical text alignment style rule, if set.
+func (s Style) UnsetAlignVertical() Style {
+	s.unset(alignVerticalKey)
+	return s
+}
+
+// UnsetPadding removes all padding style rules.
+func (s Style) UnsetPadding() Style {
+	s.unset(paddingLeftKey)
+	s.unset(paddingRightKey)
+	s.unset(paddingTopKey)
+	s.unset(paddingBottomKey)
+	return s
+}
+
+// UnsetPaddingLeft removes the left padding style rule, if set.
+func (s Style) UnsetPaddingLeft() Style {
+	s.unset(paddingLeftKey)
+	return s
+}
+
+// UnsetPaddingRight removes the right padding style rule, if set.
+func (s Style) UnsetPaddingRight() Style {
+	s.unset(paddingRightKey)
+	return s
+}
+
+// UnsetPaddingTop removes the top padding style rule, if set.
+func (s Style) UnsetPaddingTop() Style {
+	s.unset(paddingTopKey)
+	return s
+}
+
+// UnsetPaddingBottom removes the bottom padding style rule, if set.
+func (s Style) UnsetPaddingBottom() Style {
+	s.unset(paddingBottomKey)
+	return s
+}
+
+// UnsetColorWhitespace removes the rule for coloring padding, if set.
+func (s Style) UnsetColorWhitespace() Style {
+	s.unset(colorWhitespaceKey)
+	return s
+}
+
+// UnsetMargins removes all margin style rules.
+func (s Style) UnsetMargins() Style {
+	s.unset(marginLeftKey)
+	s.unset(marginRightKey)
+	s.unset(marginTopKey)
+	s.unset(marginBottomKey)
+	return s
+}
+
+// UnsetMarginLeft removes the left margin style rule, if set.
+func (s Style) UnsetMarginLeft() Style {
+	s.unset(marginLeftKey)
+	return s
+}
+
+// UnsetMarginRight removes the right margin style rule, if set.
+func (s Style) UnsetMarginRight() Style {
+	s.unset(marginRightKey)
+	return s
+}
+
+// UnsetMarginTop removes the top margin style rule, if set.
+func (s Style) UnsetMarginTop() Style {
+	s.unset(marginTopKey)
+	return s
+}
+
+// UnsetMarginBottom removes the bottom margin style rule, if set.
+func (s Style) UnsetMarginBottom() Style {
+	s.unset(marginBottomKey)
+	return s
+}
+
+// UnsetMarginBackground removes the margin's background color. Note that the
+// margin's background color can be set from the background color of another
+// style during inheritance.
+func (s Style) UnsetMarginBackground() Style {
+	s.unset(marginBackgroundKey)
+	return s
+}
+
+// UnsetBorderStyle removes the border style rule, if set.
+func (s Style) UnsetBorderStyle() Style {
+	s.unset(borderStyleKey)
+	return s
+}
+
+// UnsetBorderTop removes the border top style rule, if set.
+func (s Style) UnsetBorderTop() Style {
+	s.unset(borderTopKey)
+	return s
+}
+
+// UnsetBorderRight removes the border right style rule, if set.
+func (s Style) UnsetBorderRight() Style {
+	s.unset(borderRightKey)
+	return s
+}
+
+// UnsetBorderBottom removes the border bottom style rule, if set.
+func (s Style) UnsetBorderBottom() Style {
+	s.unset(borderBottomKey)
+	return s
+}
+
+// UnsetBorderLeft removes the border left style rule, if set.
+func (s Style) UnsetBorderLeft() Style {
+	s.unset(borderLeftKey)
+	return s
+}
+
+// UnsetBorderForeground removes all border foreground color styles, if set.
+func (s Style) UnsetBorderForeground() Style {
+	s.unset(borderTopForegroundKey)
+	s.unset(borderRightForegroundKey)
+	s.unset(borderBottomForegroundKey)
+	s.unset(borderLeftForegroundKey)
+	return s
+}
+
+// UnsetBorderTopForeground removes the top border foreground color rule,
+// if set.
+func (s Style) UnsetBorderTopForeground() Style {
+	s.unset(borderTopForegroundKey)
+	return s
+}
+
+// UnsetBorderRightForeground removes the right border foreground color rule,
+// if set.
+func (s Style) UnsetBorderRightForeground() Style {
+	s.unset(borderRightForegroundKey)
+	return s
+}
+
+// UnsetBorderBottomForeground removes the bottom border foreground color
+// rule, if set.
+func (s Style) UnsetBorderBottomForeground() Style {
+	s.unset(borderBottomForegroundKey)
+	return s
+}
+
+// UnsetBorderLeftForeground removes the left border foreground color rule,
+// if set.
+func (s Style) UnsetBorderLeftForeground() Style {
+	s.unset(borderLeftForegroundKey)
+	return s
+}
+
+// UnsetBorderBackground removes all border background color styles, if
+// set.
+func (s Style) UnsetBorderBackground() Style {
+	s.unset(borderTopBackgroundKey)
+	s.unset(borderRightBackgroundKey)
+	s.unset(borderBottomBackgroundKey)
+	s.unset(borderLeftBackgroundKey)
+	return s
+}
+
+// UnsetBorderTopBackgroundColor removes the top border background color rule,
+// if set.
+//
+// Deprecated: This function simply calls Style.UnsetBorderTopBackground.
+func (s Style) UnsetBorderTopBackgroundColor() Style {
+	return s.UnsetBorderTopBackground()
+}
+
+// UnsetBorderTopBackground removes the top border background color rule,
+// if set.
+func (s Style) UnsetBorderTopBackground() Style {
+	s.unset(borderTopBackgroundKey)
+	return s
+}
+
+// UnsetBorderRightBackground removes the right border background color
+// rule, if set.
+func (s Style) UnsetBorderRightBackground() Style {
+	s.unset(borderRightBackgroundKey)
+	return s
+}
+
+// UnsetBorderBottomBackground removes the bottom border background color
+// rule, if set.
+func (s Style) UnsetBorderBottomBackground() Style {
+	s.unset(borderBottomBackgroundKey)
+	return s
+}
+
+// UnsetBorderLeftBackground removes the left border color rule, if set.
+func (s Style) UnsetBorderLeftBackground() Style {
+	s.unset(borderLeftBackgroundKey)
+	return s
+}
+
+// UnsetInline removes the inline style rule, if set.
+func (s Style) UnsetInline() Style {
+	s.unset(inlineKey)
+	return s
+}
+
+// UnsetMaxWidth removes the max width style rule, if set.
+func (s Style) UnsetMaxWidth() Style {
+	s.unset(maxWidthKey)
+	return s
+}
+
+// UnsetMaxHeight removes the max height style rule, if set.
+func (s Style) UnsetMaxHeight() Style {
+	s.unset(maxHeightKey)
+	return s
+}
+
+// UnsetTabWidth removes the tab width style rule, if set.
+func (s Style) UnsetTabWidth() Style {
+	s.unset(tabWidthKey)
+	return s
+}
+
+// UnsetUnderlineSpaces removes the value set by UnderlineSpaces.
+func (s Style) UnsetUnderlineSpaces() Style {
+	s.unset(underlineSpacesKey)
+	return s
+}
+
+// UnsetStrikethroughSpaces removes the value set by StrikethroughSpaces.
+func (s Style) UnsetStrikethroughSpaces() Style {
+	s.unset(strikethroughSpacesKey)
+	return s
+}
+
+// UnsetTransform removes the value set by Transform.
+func (s Style) UnsetTransform() Style {
+	s.unset(transformKey)
+	return s
+}
+
+// UnsetString sets the underlying string value to the empty string.
+func (s Style) UnsetString() Style {
+	s.value = ""
+	return s
+}
diff --git a/vendor/github.com/charmbracelet/lipgloss/whitespace.go b/vendor/github.com/charmbracelet/lipgloss/whitespace.go
new file mode 100644
index 0000000..040dc98
--- /dev/null
+++ b/vendor/github.com/charmbracelet/lipgloss/whitespace.go
@@ -0,0 +1,83 @@
+package lipgloss
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+	"github.com/muesli/termenv"
+)
+
+// whitespace is a whitespace renderer.
+type whitespace struct {
+	re    *Renderer
+	style termenv.Style
+	chars string
+}
+
+// newWhitespace creates a new whitespace renderer. The order of the options
+// matters, if you're using WithWhitespaceRenderer, make sure it comes first as
+// other options might depend on it.
+func newWhitespace(r *Renderer, opts ...WhitespaceOption) *whitespace {
+	w := &whitespace{
+		re:    r,
+		style: r.ColorProfile().String(),
+	}
+	for _, opt := range opts {
+		opt(w)
+	}
+	return w
+}
+
+// Render whitespaces.
+func (w whitespace) render(width int) string {
+	if w.chars == "" {
+		w.chars = " "
+	}
+
+	r := []rune(w.chars)
+	j := 0
+	b := strings.Builder{}
+
+	// Cycle through runes and print them into the whitespace.
+	for i := 0; i < width; {
+		b.WriteRune(r[j])
+		j++
+		if j >= len(r) {
+			j = 0
+		}
+		i += ansi.StringWidth(string(r[j]))
+	}
+
+	// Fill any extra gaps white spaces. This might be necessary if any runes
+	// are more than one cell wide, which could leave a one-rune gap.
+	short := width - ansi.StringWidth(b.String())
+	if short > 0 {
+		b.WriteString(strings.Repeat(" ", short))
+	}
+
+	return w.style.Styled(b.String())
+}
+
+// WhitespaceOption sets a styling rule for rendering whitespace.
+type WhitespaceOption func(*whitespace)
+
+// WithWhitespaceForeground sets the color of the characters in the whitespace.
+func WithWhitespaceForeground(c TerminalColor) WhitespaceOption {
+	return func(w *whitespace) {
+		w.style = w.style.Foreground(c.color(w.re))
+	}
+}
+
+// WithWhitespaceBackground sets the background color of the whitespace.
+func WithWhitespaceBackground(c TerminalColor) WhitespaceOption {
+	return func(w *whitespace) {
+		w.style = w.style.Background(c.color(w.re))
+	}
+}
+
+// WithWhitespaceChars sets the characters to be rendered in the whitespace.
+func WithWhitespaceChars(s string) WhitespaceOption {
+	return func(w *whitespace) {
+		w.chars = s
+	}
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/LICENSE b/vendor/github.com/charmbracelet/x/ansi/LICENSE
new file mode 100644
index 0000000..65a5654
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2023 Charmbracelet, Inc.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/vendor/github.com/charmbracelet/x/ansi/ansi.go b/vendor/github.com/charmbracelet/x/ansi/ansi.go
new file mode 100644
index 0000000..48d873c
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/ansi.go
@@ -0,0 +1,11 @@
+package ansi
+
+import "io"
+
+// Execute is a function that "execute" the given escape sequence by writing it
+// to the provided output writter.
+//
+// This is a syntactic sugar over [io.WriteString].
+func Execute(w io.Writer, s string) (int, error) {
+	return io.WriteString(w, s)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/ascii.go b/vendor/github.com/charmbracelet/x/ansi/ascii.go
new file mode 100644
index 0000000..188582f
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/ascii.go
@@ -0,0 +1,8 @@
+package ansi
+
+const (
+	// SP is the space character (Char: \x20).
+	SP = 0x20
+	// DEL is the delete character (Caret: ^?, Char: \x7f).
+	DEL = 0x7F
+)
diff --git a/vendor/github.com/charmbracelet/x/ansi/background.go b/vendor/github.com/charmbracelet/x/ansi/background.go
new file mode 100644
index 0000000..2383cf0
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/background.go
@@ -0,0 +1,169 @@
+package ansi
+
+import (
+	"fmt"
+	"image/color"
+)
+
+// Colorizer is a [color.Color] interface that can be formatted as a string.
+type Colorizer interface {
+	color.Color
+	fmt.Stringer
+}
+
+// HexColorizer is a [color.Color] that can be formatted as a hex string.
+type HexColorizer struct{ color.Color }
+
+var _ Colorizer = HexColorizer{}
+
+// String returns the color as a hex string. If the color is nil, an empty
+// string is returned.
+func (h HexColorizer) String() string {
+	if h.Color == nil {
+		return ""
+	}
+	r, g, b, _ := h.RGBA()
+	// Get the lower 8 bits
+	r &= 0xff
+	g &= 0xff
+	b &= 0xff
+	return fmt.Sprintf("#%02x%02x%02x", uint8(r), uint8(g), uint8(b)) //nolint:gosec
+}
+
+// XRGBColorizer is a [color.Color] that can be formatted as an XParseColor
+// rgb: string.
+//
+// See: https://linux.die.net/man/3/xparsecolor
+type XRGBColorizer struct{ color.Color }
+
+var _ Colorizer = XRGBColorizer{}
+
+// String returns the color as an XParseColor rgb: string. If the color is nil,
+// an empty string is returned.
+func (x XRGBColorizer) String() string {
+	if x.Color == nil {
+		return ""
+	}
+	r, g, b, _ := x.RGBA()
+	// Get the lower 8 bits
+	return fmt.Sprintf("rgb:%04x/%04x/%04x", r, g, b)
+}
+
+// XRGBAColorizer is a [color.Color] that can be formatted as an XParseColor
+// rgba: string.
+//
+// See: https://linux.die.net/man/3/xparsecolor
+type XRGBAColorizer struct{ color.Color }
+
+var _ Colorizer = XRGBAColorizer{}
+
+// String returns the color as an XParseColor rgba: string. If the color is nil,
+// an empty string is returned.
+func (x XRGBAColorizer) String() string {
+	if x.Color == nil {
+		return ""
+	}
+	r, g, b, a := x.RGBA()
+	// Get the lower 8 bits
+	return fmt.Sprintf("rgba:%04x/%04x/%04x/%04x", r, g, b, a)
+}
+
+// SetForegroundColor returns a sequence that sets the default terminal
+// foreground color.
+//
+//	OSC 10 ; color ST
+//	OSC 10 ; color BEL
+//
+// Where color is the encoded color number.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func SetForegroundColor(c color.Color) string {
+	var s string
+	switch c := c.(type) {
+	case Colorizer:
+		s = c.String()
+	case fmt.Stringer:
+		s = c.String()
+	default:
+		s = HexColorizer{c}.String()
+	}
+	return "\x1b]10;" + s + "\x07"
+}
+
+// RequestForegroundColor is a sequence that requests the current default
+// terminal foreground color.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+const RequestForegroundColor = "\x1b]10;?\x07"
+
+// ResetForegroundColor is a sequence that resets the default terminal
+// foreground color.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+const ResetForegroundColor = "\x1b]110\x07"
+
+// SetBackgroundColor returns a sequence that sets the default terminal
+// background color.
+//
+//	OSC 11 ; color ST
+//	OSC 11 ; color BEL
+//
+// Where color is the encoded color number.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func SetBackgroundColor(c color.Color) string {
+	var s string
+	switch c := c.(type) {
+	case Colorizer:
+		s = c.String()
+	case fmt.Stringer:
+		s = c.String()
+	default:
+		s = HexColorizer{c}.String()
+	}
+	return "\x1b]11;" + s + "\x07"
+}
+
+// RequestBackgroundColor is a sequence that requests the current default
+// terminal background color.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+const RequestBackgroundColor = "\x1b]11;?\x07"
+
+// ResetBackgroundColor is a sequence that resets the default terminal
+// background color.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+const ResetBackgroundColor = "\x1b]111\x07"
+
+// SetCursorColor returns a sequence that sets the terminal cursor color.
+//
+//	OSC 12 ; color ST
+//	OSC 12 ; color BEL
+//
+// Where color is the encoded color number.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func SetCursorColor(c color.Color) string {
+	var s string
+	switch c := c.(type) {
+	case Colorizer:
+		s = c.String()
+	case fmt.Stringer:
+		s = c.String()
+	default:
+		s = HexColorizer{c}.String()
+	}
+	return "\x1b]12;" + s + "\x07"
+}
+
+// RequestCursorColor is a sequence that requests the current terminal cursor
+// color.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+const RequestCursorColor = "\x1b]12;?\x07"
+
+// ResetCursorColor is a sequence that resets the terminal cursor color.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+const ResetCursorColor = "\x1b]112\x07"
diff --git a/vendor/github.com/charmbracelet/x/ansi/c0.go b/vendor/github.com/charmbracelet/x/ansi/c0.go
new file mode 100644
index 0000000..28ff7c2
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/c0.go
@@ -0,0 +1,79 @@
+package ansi
+
+// C0 control characters.
+//
+// These range from (0x00-0x1F) as defined in ISO 646 (ASCII).
+// See: https://en.wikipedia.org/wiki/C0_and_C1_control_codes
+const (
+	// NUL is the null character (Caret: ^@, Char: \0).
+	NUL = 0x00
+	// SOH is the start of heading character (Caret: ^A).
+	SOH = 0x01
+	// STX is the start of text character (Caret: ^B).
+	STX = 0x02
+	// ETX is the end of text character (Caret: ^C).
+	ETX = 0x03
+	// EOT is the end of transmission character (Caret: ^D).
+	EOT = 0x04
+	// ENQ is the enquiry character (Caret: ^E).
+	ENQ = 0x05
+	// ACK is the acknowledge character (Caret: ^F).
+	ACK = 0x06
+	// BEL is the bell character (Caret: ^G, Char: \a).
+	BEL = 0x07
+	// BS is the backspace character (Caret: ^H, Char: \b).
+	BS = 0x08
+	// HT is the horizontal tab character (Caret: ^I, Char: \t).
+	HT = 0x09
+	// LF is the line feed character (Caret: ^J, Char: \n).
+	LF = 0x0A
+	// VT is the vertical tab character (Caret: ^K, Char: \v).
+	VT = 0x0B
+	// FF is the form feed character (Caret: ^L, Char: \f).
+	FF = 0x0C
+	// CR is the carriage return character (Caret: ^M, Char: \r).
+	CR = 0x0D
+	// SO is the shift out character (Caret: ^N).
+	SO = 0x0E
+	// SI is the shift in character (Caret: ^O).
+	SI = 0x0F
+	// DLE is the data link escape character (Caret: ^P).
+	DLE = 0x10
+	// DC1 is the device control 1 character (Caret: ^Q).
+	DC1 = 0x11
+	// DC2 is the device control 2 character (Caret: ^R).
+	DC2 = 0x12
+	// DC3 is the device control 3 character (Caret: ^S).
+	DC3 = 0x13
+	// DC4 is the device control 4 character (Caret: ^T).
+	DC4 = 0x14
+	// NAK is the negative acknowledge character (Caret: ^U).
+	NAK = 0x15
+	// SYN is the synchronous idle character (Caret: ^V).
+	SYN = 0x16
+	// ETB is the end of transmission block character (Caret: ^W).
+	ETB = 0x17
+	// CAN is the cancel character (Caret: ^X).
+	CAN = 0x18
+	// EM is the end of medium character (Caret: ^Y).
+	EM = 0x19
+	// SUB is the substitute character (Caret: ^Z).
+	SUB = 0x1A
+	// ESC is the escape character (Caret: ^[, Char: \e).
+	ESC = 0x1B
+	// FS is the file separator character (Caret: ^\).
+	FS = 0x1C
+	// GS is the group separator character (Caret: ^]).
+	GS = 0x1D
+	// RS is the record separator character (Caret: ^^).
+	RS = 0x1E
+	// US is the unit separator character (Caret: ^_).
+	US = 0x1F
+
+	// LS0 is the locking shift 0 character.
+	// This is an alias for [SI].
+	LS0 = SI
+	// LS1 is the locking shift 1 character.
+	// This is an alias for [SO].
+	LS1 = SO
+)
diff --git a/vendor/github.com/charmbracelet/x/ansi/c1.go b/vendor/github.com/charmbracelet/x/ansi/c1.go
new file mode 100644
index 0000000..71058f5
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/c1.go
@@ -0,0 +1,72 @@
+package ansi
+
+// C1 control characters.
+//
+// These range from (0x80-0x9F) as defined in ISO 6429 (ECMA-48).
+// See: https://en.wikipedia.org/wiki/C0_and_C1_control_codes
+const (
+	// PAD is the padding character.
+	PAD = 0x80
+	// HOP is the high octet preset character.
+	HOP = 0x81
+	// BPH is the break permitted here character.
+	BPH = 0x82
+	// NBH is the no break here character.
+	NBH = 0x83
+	// IND is the index character.
+	IND = 0x84
+	// NEL is the next line character.
+	NEL = 0x85
+	// SSA is the start of selected area character.
+	SSA = 0x86
+	// ESA is the end of selected area character.
+	ESA = 0x87
+	// HTS is the horizontal tab set character.
+	HTS = 0x88
+	// HTJ is the horizontal tab with justification character.
+	HTJ = 0x89
+	// VTS is the vertical tab set character.
+	VTS = 0x8A
+	// PLD is the partial line forward character.
+	PLD = 0x8B
+	// PLU is the partial line backward character.
+	PLU = 0x8C
+	// RI is the reverse index character.
+	RI = 0x8D
+	// SS2 is the single shift 2 character.
+	SS2 = 0x8E
+	// SS3 is the single shift 3 character.
+	SS3 = 0x8F
+	// DCS is the device control string character.
+	DCS = 0x90
+	// PU1 is the private use 1 character.
+	PU1 = 0x91
+	// PU2 is the private use 2 character.
+	PU2 = 0x92
+	// STS is the set transmit state character.
+	STS = 0x93
+	// CCH is the cancel character.
+	CCH = 0x94
+	// MW is the message waiting character.
+	MW = 0x95
+	// SPA is the start of guarded area character.
+	SPA = 0x96
+	// EPA is the end of guarded area character.
+	EPA = 0x97
+	// SOS is the start of string character.
+	SOS = 0x98
+	// SGCI is the single graphic character introducer character.
+	SGCI = 0x99
+	// SCI is the single character introducer character.
+	SCI = 0x9A
+	// CSI is the control sequence introducer character.
+	CSI = 0x9B
+	// ST is the string terminator character.
+	ST = 0x9C
+	// OSC is the operating system command character.
+	OSC = 0x9D
+	// PM is the privacy message character.
+	PM = 0x9E
+	// APC is the application program command character.
+	APC = 0x9F
+)
diff --git a/vendor/github.com/charmbracelet/x/ansi/charset.go b/vendor/github.com/charmbracelet/x/ansi/charset.go
new file mode 100644
index 0000000..50fff51
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/charset.go
@@ -0,0 +1,55 @@
+package ansi
+
+// SelectCharacterSet sets the G-set character designator to the specified
+// character set.
+//
+//	ESC Ps Pd
+//
+// Where Ps is the G-set character designator, and Pd is the identifier.
+// For 94-character sets, the designator can be one of:
+//   - ( G0
+//   - ) G1
+//   - * G2
+//   - + G3
+//
+// For 96-character sets, the designator can be one of:
+//   - - G1
+//   - . G2
+//   - / G3
+//
+// Some common 94-character sets are:
+//   - 0 DEC Special Drawing Set
+//   - A United Kingdom (UK)
+//   - B United States (USASCII)
+//
+// Examples:
+//
+//	ESC ( B  Select character set G0 = United States (USASCII)
+//	ESC ( 0  Select character set G0 = Special Character and Line Drawing Set
+//	ESC ) 0  Select character set G1 = Special Character and Line Drawing Set
+//	ESC * A  Select character set G2 = United Kingdom (UK)
+//
+// See: https://vt100.net/docs/vt510-rm/SCS.html
+func SelectCharacterSet(gset byte, charset byte) string {
+	return "\x1b" + string(gset) + string(charset)
+}
+
+// SCS is an alias for SelectCharacterSet.
+func SCS(gset byte, charset byte) string {
+	return SelectCharacterSet(gset, charset)
+}
+
+// Locking Shift 1 Right (LS1R) shifts G1 into GR character set.
+const LS1R = "\x1b~"
+
+// Locking Shift 2 (LS2) shifts G2 into GL character set.
+const LS2 = "\x1bn"
+
+// Locking Shift 2 Right (LS2R) shifts G2 into GR character set.
+const LS2R = "\x1b}"
+
+// Locking Shift 3 (LS3) shifts G3 into GL character set.
+const LS3 = "\x1bo"
+
+// Locking Shift 3 Right (LS3R) shifts G3 into GR character set.
+const LS3R = "\x1b|"
diff --git a/vendor/github.com/charmbracelet/x/ansi/clipboard.go b/vendor/github.com/charmbracelet/x/ansi/clipboard.go
new file mode 100644
index 0000000..94d26c3
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/clipboard.go
@@ -0,0 +1,75 @@
+package ansi
+
+import "encoding/base64"
+
+// Clipboard names.
+const (
+	SystemClipboard  = 'c'
+	PrimaryClipboard = 'p'
+)
+
+// SetClipboard returns a sequence for manipulating the clipboard.
+//
+//	OSC 52 ; Pc ; Pd ST
+//	OSC 52 ; Pc ; Pd BEL
+//
+// Where Pc is the clipboard name and Pd is the base64 encoded data.
+// Empty data or invalid base64 data will reset the clipboard.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func SetClipboard(c byte, d string) string {
+	if d != "" {
+		d = base64.StdEncoding.EncodeToString([]byte(d))
+	}
+	return "\x1b]52;" + string(c) + ";" + d + "\x07"
+}
+
+// SetSystemClipboard returns a sequence for setting the system clipboard.
+//
+// This is equivalent to SetClipboard(SystemClipboard, d).
+func SetSystemClipboard(d string) string {
+	return SetClipboard(SystemClipboard, d)
+}
+
+// SetPrimaryClipboard returns a sequence for setting the primary clipboard.
+//
+// This is equivalent to SetClipboard(PrimaryClipboard, d).
+func SetPrimaryClipboard(d string) string {
+	return SetClipboard(PrimaryClipboard, d)
+}
+
+// ResetClipboard returns a sequence for resetting the clipboard.
+//
+// This is equivalent to SetClipboard(c, "").
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func ResetClipboard(c byte) string {
+	return SetClipboard(c, "")
+}
+
+// ResetSystemClipboard is a sequence for resetting the system clipboard.
+//
+// This is equivalent to ResetClipboard(SystemClipboard).
+const ResetSystemClipboard = "\x1b]52;c;\x07"
+
+// ResetPrimaryClipboard is a sequence for resetting the primary clipboard.
+//
+// This is equivalent to ResetClipboard(PrimaryClipboard).
+const ResetPrimaryClipboard = "\x1b]52;p;\x07"
+
+// RequestClipboard returns a sequence for requesting the clipboard.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func RequestClipboard(c byte) string {
+	return "\x1b]52;" + string(c) + ";?\x07"
+}
+
+// RequestSystemClipboard is a sequence for requesting the system clipboard.
+//
+// This is equivalent to RequestClipboard(SystemClipboard).
+const RequestSystemClipboard = "\x1b]52;c;?\x07"
+
+// RequestPrimaryClipboard is a sequence for requesting the primary clipboard.
+//
+// This is equivalent to RequestClipboard(PrimaryClipboard).
+const RequestPrimaryClipboard = "\x1b]52;p;?\x07"
diff --git a/vendor/github.com/charmbracelet/x/ansi/color.go b/vendor/github.com/charmbracelet/x/ansi/color.go
new file mode 100644
index 0000000..77f8a08
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/color.go
@@ -0,0 +1,196 @@
+package ansi
+
+import (
+	"image/color"
+)
+
+// Technically speaking, the 16 basic ANSI colors are arbitrary and can be
+// customized at the terminal level. Given that, we're returning what we feel
+// are good defaults.
+//
+// This could also be a slice, but we use a map to make the mappings very
+// explicit.
+//
+// See: https://www.ditig.com/publications/256-colors-cheat-sheet
+var lowANSI = map[uint32]uint32{
+	0:  0x000000, // black
+	1:  0x800000, // red
+	2:  0x008000, // green
+	3:  0x808000, // yellow
+	4:  0x000080, // blue
+	5:  0x800080, // magenta
+	6:  0x008080, // cyan
+	7:  0xc0c0c0, // white
+	8:  0x808080, // bright black
+	9:  0xff0000, // bright red
+	10: 0x00ff00, // bright green
+	11: 0xffff00, // bright yellow
+	12: 0x0000ff, // bright blue
+	13: 0xff00ff, // bright magenta
+	14: 0x00ffff, // bright cyan
+	15: 0xffffff, // bright white
+}
+
+// Color is a color that can be used in a terminal. ANSI (including
+// ANSI256) and 24-bit "true colors" fall under this category.
+type Color interface {
+	color.Color
+}
+
+// BasicColor is an ANSI 3-bit or 4-bit color with a value from 0 to 15.
+type BasicColor uint8
+
+var _ Color = BasicColor(0)
+
+const (
+	// Black is the ANSI black color.
+	Black BasicColor = iota
+
+	// Red is the ANSI red color.
+	Red
+
+	// Green is the ANSI green color.
+	Green
+
+	// Yellow is the ANSI yellow color.
+	Yellow
+
+	// Blue is the ANSI blue color.
+	Blue
+
+	// Magenta is the ANSI magenta color.
+	Magenta
+
+	// Cyan is the ANSI cyan color.
+	Cyan
+
+	// White is the ANSI white color.
+	White
+
+	// BrightBlack is the ANSI bright black color.
+	BrightBlack
+
+	// BrightRed is the ANSI bright red color.
+	BrightRed
+
+	// BrightGreen is the ANSI bright green color.
+	BrightGreen
+
+	// BrightYellow is the ANSI bright yellow color.
+	BrightYellow
+
+	// BrightBlue is the ANSI bright blue color.
+	BrightBlue
+
+	// BrightMagenta is the ANSI bright magenta color.
+	BrightMagenta
+
+	// BrightCyan is the ANSI bright cyan color.
+	BrightCyan
+
+	// BrightWhite is the ANSI bright white color.
+	BrightWhite
+)
+
+// RGBA returns the red, green, blue and alpha components of the color. It
+// satisfies the color.Color interface.
+func (c BasicColor) RGBA() (uint32, uint32, uint32, uint32) {
+	ansi := uint32(c)
+	if ansi > 15 {
+		return 0, 0, 0, 0xffff
+	}
+
+	r, g, b := ansiToRGB(ansi)
+	return toRGBA(r, g, b)
+}
+
+// ExtendedColor is an ANSI 256 (8-bit) color with a value from 0 to 255.
+type ExtendedColor uint8
+
+var _ Color = ExtendedColor(0)
+
+// RGBA returns the red, green, blue and alpha components of the color. It
+// satisfies the color.Color interface.
+func (c ExtendedColor) RGBA() (uint32, uint32, uint32, uint32) {
+	r, g, b := ansiToRGB(uint32(c))
+	return toRGBA(r, g, b)
+}
+
+// TrueColor is a 24-bit color that can be used in the terminal.
+// This can be used to represent RGB colors.
+//
+// For example, the color red can be represented as:
+//
+//	TrueColor(0xff0000)
+type TrueColor uint32
+
+var _ Color = TrueColor(0)
+
+// RGBA returns the red, green, blue and alpha components of the color. It
+// satisfies the color.Color interface.
+func (c TrueColor) RGBA() (uint32, uint32, uint32, uint32) {
+	r, g, b := hexToRGB(uint32(c))
+	return toRGBA(r, g, b)
+}
+
+// ansiToRGB converts an ANSI color to a 24-bit RGB color.
+//
+//	r, g, b := ansiToRGB(57)
+func ansiToRGB(ansi uint32) (uint32, uint32, uint32) {
+	// For out-of-range values return black.
+	if ansi > 255 {
+		return 0, 0, 0
+	}
+
+	// Low ANSI.
+	if ansi < 16 {
+		h, ok := lowANSI[ansi]
+		if !ok {
+			return 0, 0, 0
+		}
+		r, g, b := hexToRGB(h)
+		return r, g, b
+	}
+
+	// Grays.
+	if ansi > 231 {
+		s := (ansi-232)*10 + 8
+		return s, s, s
+	}
+
+	// ANSI256.
+	n := ansi - 16
+	b := n % 6
+	g := (n - b) / 6 % 6
+	r := (n - b - g*6) / 36 % 6
+	for _, v := range []*uint32{&r, &g, &b} {
+		if *v > 0 {
+			c := *v*40 + 55
+			*v = c
+		}
+	}
+
+	return r, g, b
+}
+
+// hexToRGB converts a number in hexadecimal format to red, green, and blue
+// values.
+//
+//	r, g, b := hexToRGB(0x0000FF)
+func hexToRGB(hex uint32) (uint32, uint32, uint32) {
+	return hex >> 16 & 0xff, hex >> 8 & 0xff, hex & 0xff
+}
+
+// toRGBA converts an RGB 8-bit color values to 32-bit color values suitable
+// for color.Color.
+//
+// color.Color requires 16-bit color values, so we duplicate the 8-bit values
+// to fill the 16-bit values.
+//
+// This always returns 0xffff (opaque) for the alpha channel.
+func toRGBA(r, g, b uint32) (uint32, uint32, uint32, uint32) {
+	r |= r << 8
+	g |= g << 8
+	b |= b << 8
+	return r, g, b, 0xffff
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/ctrl.go b/vendor/github.com/charmbracelet/x/ansi/ctrl.go
new file mode 100644
index 0000000..8ca744c
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/ctrl.go
@@ -0,0 +1,137 @@
+package ansi
+
+import (
+	"strconv"
+	"strings"
+)
+
+// RequestNameVersion (XTVERSION) is a control sequence that requests the
+// terminal's name and version. It responds with a DSR sequence identifying the
+// terminal.
+//
+//	CSI > 0 q
+//	DCS > | text ST
+//
+// See https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-PC-Style-Function-Keys
+const (
+	RequestNameVersion = "\x1b[>q"
+	XTVERSION          = RequestNameVersion
+)
+
+// RequestXTVersion is a control sequence that requests the terminal's XTVERSION. It responds with a DSR sequence identifying the version.
+//
+//	CSI > Ps q
+//	DCS > | text ST
+//
+// See https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-PC-Style-Function-Keys
+//
+// Deprecated: use [RequestNameVersion] instead.
+const RequestXTVersion = RequestNameVersion
+
+// PrimaryDeviceAttributes (DA1) is a control sequence that reports the
+// terminal's primary device attributes.
+//
+//	CSI c
+//	CSI 0 c
+//	CSI ? Ps ; ... c
+//
+// If no attributes are given, or if the attribute is 0, this function returns
+// the request sequence. Otherwise, it returns the response sequence.
+//
+// See https://vt100.net/docs/vt510-rm/DA1.html
+func PrimaryDeviceAttributes(attrs ...int) string {
+	if len(attrs) == 0 {
+		return RequestPrimaryDeviceAttributes
+	} else if len(attrs) == 1 && attrs[0] == 0 {
+		return "\x1b[0c"
+	}
+
+	as := make([]string, len(attrs))
+	for i, a := range attrs {
+		as[i] = strconv.Itoa(a)
+	}
+	return "\x1b[?" + strings.Join(as, ";") + "c"
+}
+
+// DA1 is an alias for [PrimaryDeviceAttributes].
+func DA1(attrs ...int) string {
+	return PrimaryDeviceAttributes(attrs...)
+}
+
+// RequestPrimaryDeviceAttributes is a control sequence that requests the
+// terminal's primary device attributes (DA1).
+//
+//	CSI c
+//
+// See https://vt100.net/docs/vt510-rm/DA1.html
+const RequestPrimaryDeviceAttributes = "\x1b[c"
+
+// SecondaryDeviceAttributes (DA2) is a control sequence that reports the
+// terminal's secondary device attributes.
+//
+//	CSI > c
+//	CSI > 0 c
+//	CSI > Ps ; ... c
+//
+// See https://vt100.net/docs/vt510-rm/DA2.html
+func SecondaryDeviceAttributes(attrs ...int) string {
+	if len(attrs) == 0 {
+		return RequestSecondaryDeviceAttributes
+	}
+
+	as := make([]string, len(attrs))
+	for i, a := range attrs {
+		as[i] = strconv.Itoa(a)
+	}
+	return "\x1b[>" + strings.Join(as, ";") + "c"
+}
+
+// DA2 is an alias for [SecondaryDeviceAttributes].
+func DA2(attrs ...int) string {
+	return SecondaryDeviceAttributes(attrs...)
+}
+
+// RequestSecondaryDeviceAttributes is a control sequence that requests the
+// terminal's secondary device attributes (DA2).
+//
+//	CSI > c
+//
+// See https://vt100.net/docs/vt510-rm/DA2.html
+const RequestSecondaryDeviceAttributes = "\x1b[>c"
+
+// TertiaryDeviceAttributes (DA3) is a control sequence that reports the
+// terminal's tertiary device attributes.
+//
+//	CSI = c
+//	CSI = 0 c
+//	DCS ! | Text ST
+//
+// Where Text is the unit ID for the terminal.
+//
+// If no unit ID is given, or if the unit ID is 0, this function returns the
+// request sequence. Otherwise, it returns the response sequence.
+//
+// See https://vt100.net/docs/vt510-rm/DA3.html
+func TertiaryDeviceAttributes(unitID string) string {
+	switch unitID {
+	case "":
+		return RequestTertiaryDeviceAttributes
+	case "0":
+		return "\x1b[=0c"
+	}
+
+	return "\x1bP!|" + unitID + "\x1b\\"
+}
+
+// DA3 is an alias for [TertiaryDeviceAttributes].
+func DA3(unitID string) string {
+	return TertiaryDeviceAttributes(unitID)
+}
+
+// RequestTertiaryDeviceAttributes is a control sequence that requests the
+// terminal's tertiary device attributes (DA3).
+//
+//	CSI = c
+//
+// See https://vt100.net/docs/vt510-rm/DA3.html
+const RequestTertiaryDeviceAttributes = "\x1b[=c"
diff --git a/vendor/github.com/charmbracelet/x/ansi/cursor.go b/vendor/github.com/charmbracelet/x/ansi/cursor.go
new file mode 100644
index 0000000..0c364d6
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/cursor.go
@@ -0,0 +1,633 @@
+package ansi
+
+import "strconv"
+
+// SaveCursor (DECSC) is an escape sequence that saves the current cursor
+// position.
+//
+//	ESC 7
+//
+// See: https://vt100.net/docs/vt510-rm/DECSC.html
+const (
+	SaveCursor = "\x1b7"
+	DECSC      = SaveCursor
+)
+
+// RestoreCursor (DECRC) is an escape sequence that restores the cursor
+// position.
+//
+//	ESC 8
+//
+// See: https://vt100.net/docs/vt510-rm/DECRC.html
+const (
+	RestoreCursor = "\x1b8"
+	DECRC         = RestoreCursor
+)
+
+// RequestCursorPosition is an escape sequence that requests the current cursor
+// position.
+//
+//	CSI 6 n
+//
+// The terminal will report the cursor position as a CSI sequence in the
+// following format:
+//
+//	CSI Pl ; Pc R
+//
+// Where Pl is the line number and Pc is the column number.
+// See: https://vt100.net/docs/vt510-rm/CPR.html
+//
+// Deprecated: use [RequestCursorPositionReport] instead.
+const RequestCursorPosition = "\x1b[6n"
+
+// RequestExtendedCursorPosition (DECXCPR) is a sequence for requesting the
+// cursor position report including the current page number.
+//
+//	CSI ? 6 n
+//
+// The terminal will report the cursor position as a CSI sequence in the
+// following format:
+//
+//	CSI ? Pl ; Pc ; Pp R
+//
+// Where Pl is the line number, Pc is the column number, and Pp is the page
+// number.
+// See: https://vt100.net/docs/vt510-rm/DECXCPR.html
+//
+// Deprecated: use [RequestExtendedCursorPositionReport] instead.
+const RequestExtendedCursorPosition = "\x1b[?6n"
+
+// CursorUp (CUU) returns a sequence for moving the cursor up n cells.
+//
+//	CSI n A
+//
+// See: https://vt100.net/docs/vt510-rm/CUU.html
+func CursorUp(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "A"
+}
+
+// CUU is an alias for [CursorUp].
+func CUU(n int) string {
+	return CursorUp(n)
+}
+
+// CUU1 is a sequence for moving the cursor up one cell.
+const CUU1 = "\x1b[A"
+
+// CursorUp1 is a sequence for moving the cursor up one cell.
+//
+// This is equivalent to CursorUp(1).
+//
+// Deprecated: use [CUU1] instead.
+const CursorUp1 = "\x1b[A"
+
+// CursorDown (CUD) returns a sequence for moving the cursor down n cells.
+//
+//	CSI n B
+//
+// See: https://vt100.net/docs/vt510-rm/CUD.html
+func CursorDown(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "B"
+}
+
+// CUD is an alias for [CursorDown].
+func CUD(n int) string {
+	return CursorDown(n)
+}
+
+// CUD1 is a sequence for moving the cursor down one cell.
+const CUD1 = "\x1b[B"
+
+// CursorDown1 is a sequence for moving the cursor down one cell.
+//
+// This is equivalent to CursorDown(1).
+//
+// Deprecated: use [CUD1] instead.
+const CursorDown1 = "\x1b[B"
+
+// CursorForward (CUF) returns a sequence for moving the cursor right n cells.
+//
+// # CSI n C
+//
+// See: https://vt100.net/docs/vt510-rm/CUF.html
+func CursorForward(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "C"
+}
+
+// CUF is an alias for [CursorForward].
+func CUF(n int) string {
+	return CursorForward(n)
+}
+
+// CUF1 is a sequence for moving the cursor right one cell.
+const CUF1 = "\x1b[C"
+
+// CursorRight (CUF) returns a sequence for moving the cursor right n cells.
+//
+//	CSI n C
+//
+// See: https://vt100.net/docs/vt510-rm/CUF.html
+//
+// Deprecated: use [CursorForward] instead.
+func CursorRight(n int) string {
+	return CursorForward(n)
+}
+
+// CursorRight1 is a sequence for moving the cursor right one cell.
+//
+// This is equivalent to CursorRight(1).
+//
+// Deprecated: use [CUF1] instead.
+const CursorRight1 = CUF1
+
+// CursorBackward (CUB) returns a sequence for moving the cursor left n cells.
+//
+// # CSI n D
+//
+// See: https://vt100.net/docs/vt510-rm/CUB.html
+func CursorBackward(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "D"
+}
+
+// CUB is an alias for [CursorBackward].
+func CUB(n int) string {
+	return CursorBackward(n)
+}
+
+// CUB1 is a sequence for moving the cursor left one cell.
+const CUB1 = "\x1b[D"
+
+// CursorLeft (CUB) returns a sequence for moving the cursor left n cells.
+//
+//	CSI n D
+//
+// See: https://vt100.net/docs/vt510-rm/CUB.html
+//
+// Deprecated: use [CursorBackward] instead.
+func CursorLeft(n int) string {
+	return CursorBackward(n)
+}
+
+// CursorLeft1 is a sequence for moving the cursor left one cell.
+//
+// This is equivalent to CursorLeft(1).
+//
+// Deprecated: use [CUB1] instead.
+const CursorLeft1 = CUB1
+
+// CursorNextLine (CNL) returns a sequence for moving the cursor to the
+// beginning of the next line n times.
+//
+//	CSI n E
+//
+// See: https://vt100.net/docs/vt510-rm/CNL.html
+func CursorNextLine(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "E"
+}
+
+// CNL is an alias for [CursorNextLine].
+func CNL(n int) string {
+	return CursorNextLine(n)
+}
+
+// CursorPreviousLine (CPL) returns a sequence for moving the cursor to the
+// beginning of the previous line n times.
+//
+//	CSI n F
+//
+// See: https://vt100.net/docs/vt510-rm/CPL.html
+func CursorPreviousLine(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "F"
+}
+
+// CPL is an alias for [CursorPreviousLine].
+func CPL(n int) string {
+	return CursorPreviousLine(n)
+}
+
+// CursorHorizontalAbsolute (CHA) returns a sequence for moving the cursor to
+// the given column.
+//
+// Default is 1.
+//
+//	CSI n G
+//
+// See: https://vt100.net/docs/vt510-rm/CHA.html
+func CursorHorizontalAbsolute(col int) string {
+	var s string
+	if col > 0 {
+		s = strconv.Itoa(col)
+	}
+	return "\x1b[" + s + "G"
+}
+
+// CHA is an alias for [CursorHorizontalAbsolute].
+func CHA(col int) string {
+	return CursorHorizontalAbsolute(col)
+}
+
+// CursorPosition (CUP) returns a sequence for setting the cursor to the
+// given row and column.
+//
+// Default is 1,1.
+//
+//	CSI n ; m H
+//
+// See: https://vt100.net/docs/vt510-rm/CUP.html
+func CursorPosition(col, row int) string {
+	if row <= 0 && col <= 0 {
+		return HomeCursorPosition
+	}
+
+	var r, c string
+	if row > 0 {
+		r = strconv.Itoa(row)
+	}
+	if col > 0 {
+		c = strconv.Itoa(col)
+	}
+	return "\x1b[" + r + ";" + c + "H"
+}
+
+// CUP is an alias for [CursorPosition].
+func CUP(col, row int) string {
+	return CursorPosition(col, row)
+}
+
+// CursorHomePosition is a sequence for moving the cursor to the upper left
+// corner of the scrolling region. This is equivalent to `CursorPosition(1, 1)`.
+const CursorHomePosition = "\x1b[H"
+
+// SetCursorPosition (CUP) returns a sequence for setting the cursor to the
+// given row and column.
+//
+//	CSI n ; m H
+//
+// See: https://vt100.net/docs/vt510-rm/CUP.html
+//
+// Deprecated: use [CursorPosition] instead.
+func SetCursorPosition(col, row int) string {
+	if row <= 0 && col <= 0 {
+		return HomeCursorPosition
+	}
+
+	var r, c string
+	if row > 0 {
+		r = strconv.Itoa(row)
+	}
+	if col > 0 {
+		c = strconv.Itoa(col)
+	}
+	return "\x1b[" + r + ";" + c + "H"
+}
+
+// HomeCursorPosition is a sequence for moving the cursor to the upper left
+// corner of the scrolling region. This is equivalent to `SetCursorPosition(1, 1)`.
+//
+// Deprecated: use [CursorHomePosition] instead.
+const HomeCursorPosition = CursorHomePosition
+
+// MoveCursor (CUP) returns a sequence for setting the cursor to the
+// given row and column.
+//
+//	CSI n ; m H
+//
+// See: https://vt100.net/docs/vt510-rm/CUP.html
+//
+// Deprecated: use [CursorPosition] instead.
+func MoveCursor(col, row int) string {
+	return SetCursorPosition(col, row)
+}
+
+// CursorOrigin is a sequence for moving the cursor to the upper left corner of
+// the display. This is equivalent to `SetCursorPosition(1, 1)`.
+//
+// Deprecated: use [CursorHomePosition] instead.
+const CursorOrigin = "\x1b[1;1H"
+
+// MoveCursorOrigin is a sequence for moving the cursor to the upper left
+// corner of the display. This is equivalent to `SetCursorPosition(1, 1)`.
+//
+// Deprecated: use [CursorHomePosition] instead.
+const MoveCursorOrigin = CursorOrigin
+
+// CursorHorizontalForwardTab (CHT) returns a sequence for moving the cursor to
+// the next tab stop n times.
+//
+// Default is 1.
+//
+//	CSI n I
+//
+// See: https://vt100.net/docs/vt510-rm/CHT.html
+func CursorHorizontalForwardTab(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "I"
+}
+
+// CHT is an alias for [CursorHorizontalForwardTab].
+func CHT(n int) string {
+	return CursorHorizontalForwardTab(n)
+}
+
+// EraseCharacter (ECH) returns a sequence for erasing n characters and moving
+// the cursor to the right. This doesn't affect other cell attributes.
+//
+// Default is 1.
+//
+//	CSI n X
+//
+// See: https://vt100.net/docs/vt510-rm/ECH.html
+func EraseCharacter(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "X"
+}
+
+// ECH is an alias for [EraseCharacter].
+func ECH(n int) string {
+	return EraseCharacter(n)
+}
+
+// CursorBackwardTab (CBT) returns a sequence for moving the cursor to the
+// previous tab stop n times.
+//
+// Default is 1.
+//
+//	CSI n Z
+//
+// See: https://vt100.net/docs/vt510-rm/CBT.html
+func CursorBackwardTab(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "Z"
+}
+
+// CBT is an alias for [CursorBackwardTab].
+func CBT(n int) string {
+	return CursorBackwardTab(n)
+}
+
+// VerticalPositionAbsolute (VPA) returns a sequence for moving the cursor to
+// the given row.
+//
+// Default is 1.
+//
+//	CSI n d
+//
+// See: https://vt100.net/docs/vt510-rm/VPA.html
+func VerticalPositionAbsolute(row int) string {
+	var s string
+	if row > 0 {
+		s = strconv.Itoa(row)
+	}
+	return "\x1b[" + s + "d"
+}
+
+// VPA is an alias for [VerticalPositionAbsolute].
+func VPA(row int) string {
+	return VerticalPositionAbsolute(row)
+}
+
+// VerticalPositionRelative (VPR) returns a sequence for moving the cursor down
+// n rows relative to the current position.
+//
+// Default is 1.
+//
+//	CSI n e
+//
+// See: https://vt100.net/docs/vt510-rm/VPR.html
+func VerticalPositionRelative(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "e"
+}
+
+// VPR is an alias for [VerticalPositionRelative].
+func VPR(n int) string {
+	return VerticalPositionRelative(n)
+}
+
+// HorizontalVerticalPosition (HVP) returns a sequence for moving the cursor to
+// the given row and column.
+//
+// Default is 1,1.
+//
+//	CSI n ; m f
+//
+// This has the same effect as [CursorPosition].
+//
+// See: https://vt100.net/docs/vt510-rm/HVP.html
+func HorizontalVerticalPosition(col, row int) string {
+	var r, c string
+	if row > 0 {
+		r = strconv.Itoa(row)
+	}
+	if col > 0 {
+		c = strconv.Itoa(col)
+	}
+	return "\x1b[" + r + ";" + c + "f"
+}
+
+// HVP is an alias for [HorizontalVerticalPosition].
+func HVP(col, row int) string {
+	return HorizontalVerticalPosition(col, row)
+}
+
+// HorizontalVerticalHomePosition is a sequence for moving the cursor to the
+// upper left corner of the scrolling region. This is equivalent to
+// `HorizontalVerticalPosition(1, 1)`.
+const HorizontalVerticalHomePosition = "\x1b[f"
+
+// SaveCurrentCursorPosition (SCOSC) is a sequence for saving the current cursor
+// position for SCO console mode.
+//
+//	CSI s
+//
+// This acts like [DECSC], except the page number where the cursor is located
+// is not saved.
+//
+// See: https://vt100.net/docs/vt510-rm/SCOSC.html
+const (
+	SaveCurrentCursorPosition = "\x1b[s"
+	SCOSC                     = SaveCurrentCursorPosition
+)
+
+// SaveCursorPosition (SCP or SCOSC) is a sequence for saving the cursor
+// position.
+//
+//	CSI s
+//
+// This acts like Save, except the page number where the cursor is located is
+// not saved.
+//
+// See: https://vt100.net/docs/vt510-rm/SCOSC.html
+//
+// Deprecated: use [SaveCurrentCursorPosition] instead.
+const SaveCursorPosition = "\x1b[s"
+
+// RestoreCurrentCursorPosition (SCORC) is a sequence for restoring the current
+// cursor position for SCO console mode.
+//
+//	CSI u
+//
+// This acts like [DECRC], except the page number where the cursor was saved is
+// not restored.
+//
+// See: https://vt100.net/docs/vt510-rm/SCORC.html
+const (
+	RestoreCurrentCursorPosition = "\x1b[u"
+	SCORC                        = RestoreCurrentCursorPosition
+)
+
+// RestoreCursorPosition (RCP or SCORC) is a sequence for restoring the cursor
+// position.
+//
+//	CSI u
+//
+// This acts like Restore, except the cursor stays on the same page where the
+// cursor was saved.
+//
+// See: https://vt100.net/docs/vt510-rm/SCORC.html
+//
+// Deprecated: use [RestoreCurrentCursorPosition] instead.
+const RestoreCursorPosition = "\x1b[u"
+
+// SetCursorStyle (DECSCUSR) returns a sequence for changing the cursor style.
+//
+// Default is 1.
+//
+//	CSI Ps SP q
+//
+// Where Ps is the cursor style:
+//
+//	0: Blinking block
+//	1: Blinking block (default)
+//	2: Steady block
+//	3: Blinking underline
+//	4: Steady underline
+//	5: Blinking bar (xterm)
+//	6: Steady bar (xterm)
+//
+// See: https://vt100.net/docs/vt510-rm/DECSCUSR.html
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h4-Functions-using-CSI-_-ordered-by-the-final-character-lparen-s-rparen:CSI-Ps-SP-q.1D81
+func SetCursorStyle(style int) string {
+	if style < 0 {
+		style = 0
+	}
+	return "\x1b[" + strconv.Itoa(style) + " q"
+}
+
+// DECSCUSR is an alias for [SetCursorStyle].
+func DECSCUSR(style int) string {
+	return SetCursorStyle(style)
+}
+
+// SetPointerShape returns a sequence for changing the mouse pointer cursor
+// shape. Use "default" for the default pointer shape.
+//
+//	OSC 22 ; Pt ST
+//	OSC 22 ; Pt BEL
+//
+// Where Pt is the pointer shape name. The name can be anything that the
+// operating system can understand. Some common names are:
+//
+//   - copy
+//   - crosshair
+//   - default
+//   - ew-resize
+//   - n-resize
+//   - text
+//   - wait
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Operating-System-Commands
+func SetPointerShape(shape string) string {
+	return "\x1b]22;" + shape + "\x07"
+}
+
+// ReverseIndex (RI) is an escape sequence for moving the cursor up one line in
+// the same column. If the cursor is at the top margin, the screen scrolls
+// down.
+//
+// This has the same effect as [RI].
+const ReverseIndex = "\x1bM"
+
+// HorizontalPositionAbsolute (HPA) returns a sequence for moving the cursor to
+// the given column. This has the same effect as [CUP].
+//
+// Default is 1.
+//
+//	CSI n `
+//
+// See: https://vt100.net/docs/vt510-rm/HPA.html
+func HorizontalPositionAbsolute(col int) string {
+	var s string
+	if col > 0 {
+		s = strconv.Itoa(col)
+	}
+	return "\x1b[" + s + "`"
+}
+
+// HPA is an alias for [HorizontalPositionAbsolute].
+func HPA(col int) string {
+	return HorizontalPositionAbsolute(col)
+}
+
+// HorizontalPositionRelative (HPR) returns a sequence for moving the cursor
+// right n columns relative to the current position. This has the same effect
+// as [CUP].
+//
+// Default is 1.
+//
+//	CSI n a
+//
+// See: https://vt100.net/docs/vt510-rm/HPR.html
+func HorizontalPositionRelative(n int) string {
+	var s string
+	if n > 0 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "a"
+}
+
+// HPR is an alias for [HorizontalPositionRelative].
+func HPR(n int) string {
+	return HorizontalPositionRelative(n)
+}
+
+// Index (IND) is an escape sequence for moving the cursor down one line in the
+// same column. If the cursor is at the bottom margin, the screen scrolls up.
+// This has the same effect as [IND].
+const Index = "\x1bD"
diff --git a/vendor/github.com/charmbracelet/x/ansi/cwd.go b/vendor/github.com/charmbracelet/x/ansi/cwd.go
new file mode 100644
index 0000000..b03ac1b
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/cwd.go
@@ -0,0 +1,26 @@
+package ansi
+
+import (
+	"net/url"
+	"path"
+)
+
+// NotifyWorkingDirectory returns a sequence that notifies the terminal
+// of the current working directory.
+//
+//	OSC 7 ; Pt BEL
+//
+// Where Pt is a URL in the format "file://[host]/[path]".
+// Set host to "localhost" if this is a path on the local computer.
+//
+// See: https://wezfurlong.org/wezterm/shell-integration.html#osc-7-escape-sequence-to-set-the-working-directory
+// See: https://iterm2.com/documentation-escape-codes.html#:~:text=RemoteHost%20and%20CurrentDir%3A-,OSC%207,-%3B%20%5BPs%5D%20ST
+func NotifyWorkingDirectory(host string, paths ...string) string {
+	path := path.Join(paths...)
+	u := &url.URL{
+		Scheme: "file",
+		Host:   host,
+		Path:   path,
+	}
+	return "\x1b]7;" + u.String() + "\x07"
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/doc.go b/vendor/github.com/charmbracelet/x/ansi/doc.go
new file mode 100644
index 0000000..e955e9f
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/doc.go
@@ -0,0 +1,7 @@
+// Package ansi defines common ANSI escape sequences based on the ECMA-48
+// specs.
+//
+// All sequences use 7-bit C1 control codes, which are supported by most
+// terminal emulators. OSC sequences are terminated by a BEL for wider
+// compatibility with terminals.
+package ansi
diff --git a/vendor/github.com/charmbracelet/x/ansi/focus.go b/vendor/github.com/charmbracelet/x/ansi/focus.go
new file mode 100644
index 0000000..4e0207c
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/focus.go
@@ -0,0 +1,9 @@
+package ansi
+
+// Focus is an escape sequence to notify the terminal that it has focus.
+// This is used with [FocusEventMode].
+const Focus = "\x1b[I"
+
+// Blur is an escape sequence to notify the terminal that it has lost focus.
+// This is used with [FocusEventMode].
+const Blur = "\x1b[O"
diff --git a/vendor/github.com/charmbracelet/x/ansi/graphics.go b/vendor/github.com/charmbracelet/x/ansi/graphics.go
new file mode 100644
index 0000000..604fef4
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/graphics.go
@@ -0,0 +1,199 @@
+package ansi
+
+import (
+	"bytes"
+	"encoding/base64"
+	"errors"
+	"fmt"
+	"image"
+	"io"
+	"os"
+	"strings"
+
+	"github.com/charmbracelet/x/ansi/kitty"
+)
+
+// KittyGraphics returns a sequence that encodes the given image in the Kitty
+// graphics protocol.
+//
+//	APC G [comma separated options] ; [base64 encoded payload] ST
+//
+// See https://sw.kovidgoyal.net/kitty/graphics-protocol/
+func KittyGraphics(payload []byte, opts ...string) string {
+	var buf bytes.Buffer
+	buf.WriteString("\x1b_G")
+	buf.WriteString(strings.Join(opts, ","))
+	if len(payload) > 0 {
+		buf.WriteString(";")
+		buf.Write(payload)
+	}
+	buf.WriteString("\x1b\\")
+	return buf.String()
+}
+
+var (
+	// KittyGraphicsTempDir is the directory where temporary files are stored.
+	// This is used in [WriteKittyGraphics] along with [os.CreateTemp].
+	KittyGraphicsTempDir = ""
+
+	// KittyGraphicsTempPattern is the pattern used to create temporary files.
+	// This is used in [WriteKittyGraphics] along with [os.CreateTemp].
+	// The Kitty Graphics protocol requires the file path to contain the
+	// substring "tty-graphics-protocol".
+	KittyGraphicsTempPattern = "tty-graphics-protocol-*"
+)
+
+// WriteKittyGraphics writes an image using the Kitty Graphics protocol with
+// the given options to w. It chunks the written data if o.Chunk is true.
+//
+// You can omit m and use nil when rendering an image from a file. In this
+// case, you must provide a file path in o.File and use o.Transmission =
+// [kitty.File]. You can also use o.Transmission = [kitty.TempFile] to write
+// the image to a temporary file. In that case, the file path is ignored, and
+// the image is written to a temporary file that is automatically deleted by
+// the terminal.
+//
+// See https://sw.kovidgoyal.net/kitty/graphics-protocol/
+func WriteKittyGraphics(w io.Writer, m image.Image, o *kitty.Options) error {
+	if o == nil {
+		o = &kitty.Options{}
+	}
+
+	if o.Transmission == 0 && len(o.File) != 0 {
+		o.Transmission = kitty.File
+	}
+
+	var data bytes.Buffer // the data to be encoded into base64
+	e := &kitty.Encoder{
+		Compress: o.Compression == kitty.Zlib,
+		Format:   o.Format,
+	}
+
+	switch o.Transmission {
+	case kitty.Direct:
+		if err := e.Encode(&data, m); err != nil {
+			return fmt.Errorf("failed to encode direct image: %w", err)
+		}
+
+	case kitty.SharedMemory:
+		// TODO: Implement shared memory
+		return fmt.Errorf("shared memory transmission is not yet implemented")
+
+	case kitty.File:
+		if len(o.File) == 0 {
+			return kitty.ErrMissingFile
+		}
+
+		f, err := os.Open(o.File)
+		if err != nil {
+			return fmt.Errorf("failed to open file: %w", err)
+		}
+
+		defer f.Close() //nolint:errcheck
+
+		stat, err := f.Stat()
+		if err != nil {
+			return fmt.Errorf("failed to get file info: %w", err)
+		}
+
+		mode := stat.Mode()
+		if !mode.IsRegular() {
+			return fmt.Errorf("file is not a regular file")
+		}
+
+		// Write the file path to the buffer
+		if _, err := data.WriteString(f.Name()); err != nil {
+			return fmt.Errorf("failed to write file path to buffer: %w", err)
+		}
+
+	case kitty.TempFile:
+		f, err := os.CreateTemp(KittyGraphicsTempDir, KittyGraphicsTempPattern)
+		if err != nil {
+			return fmt.Errorf("failed to create file: %w", err)
+		}
+
+		defer f.Close() //nolint:errcheck
+
+		if err := e.Encode(f, m); err != nil {
+			return fmt.Errorf("failed to encode image to file: %w", err)
+		}
+
+		// Write the file path to the buffer
+		if _, err := data.WriteString(f.Name()); err != nil {
+			return fmt.Errorf("failed to write file path to buffer: %w", err)
+		}
+	}
+
+	// Encode image to base64
+	var payload bytes.Buffer // the base64 encoded image to be written to w
+	b64 := base64.NewEncoder(base64.StdEncoding, &payload)
+	if _, err := data.WriteTo(b64); err != nil {
+		return fmt.Errorf("failed to write base64 encoded image to payload: %w", err)
+	}
+	if err := b64.Close(); err != nil {
+		return err
+	}
+
+	// If not chunking, write all at once
+	if !o.Chunk {
+		_, err := io.WriteString(w, KittyGraphics(payload.Bytes(), o.Options()...))
+		return err
+	}
+
+	// Write in chunks
+	var (
+		err error
+		n   int
+	)
+	chunk := make([]byte, kitty.MaxChunkSize)
+	isFirstChunk := true
+
+	for {
+		// Stop if we read less than the chunk size [kitty.MaxChunkSize].
+		n, err = io.ReadFull(&payload, chunk)
+		if errors.Is(err, io.ErrUnexpectedEOF) || errors.Is(err, io.EOF) {
+			break
+		}
+		if err != nil {
+			return fmt.Errorf("failed to read chunk: %w", err)
+		}
+
+		opts := buildChunkOptions(o, isFirstChunk, false)
+		if _, err := io.WriteString(w, KittyGraphics(chunk[:n], opts...)); err != nil {
+			return err
+		}
+
+		isFirstChunk = false
+	}
+
+	// Write the last chunk
+	opts := buildChunkOptions(o, isFirstChunk, true)
+	_, err = io.WriteString(w, KittyGraphics(chunk[:n], opts...))
+	return err
+}
+
+// buildChunkOptions creates the options slice for a chunk
+func buildChunkOptions(o *kitty.Options, isFirstChunk, isLastChunk bool) []string {
+	var opts []string
+	if isFirstChunk {
+		opts = o.Options()
+	} else {
+		// These options are allowed in subsequent chunks
+		if o.Quite > 0 {
+			opts = append(opts, fmt.Sprintf("q=%d", o.Quite))
+		}
+		if o.Action == kitty.Frame {
+			opts = append(opts, "a=f")
+		}
+	}
+
+	if !isFirstChunk || !isLastChunk {
+		// We don't need to encode the (m=) option when we only have one chunk.
+		if isLastChunk {
+			opts = append(opts, "m=0")
+		} else {
+			opts = append(opts, "m=1")
+		}
+	}
+	return opts
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/hyperlink.go b/vendor/github.com/charmbracelet/x/ansi/hyperlink.go
new file mode 100644
index 0000000..323bfe9
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/hyperlink.go
@@ -0,0 +1,28 @@
+package ansi
+
+import "strings"
+
+// SetHyperlink returns a sequence for starting a hyperlink.
+//
+//	OSC 8 ; Params ; Uri ST
+//	OSC 8 ; Params ; Uri BEL
+//
+// To reset the hyperlink, omit the URI.
+//
+// See: https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda
+func SetHyperlink(uri string, params ...string) string {
+	var p string
+	if len(params) > 0 {
+		p = strings.Join(params, ":")
+	}
+	return "\x1b]8;" + p + ";" + uri + "\x07"
+}
+
+// ResetHyperlink returns a sequence for resetting the hyperlink.
+//
+// This is equivalent to SetHyperlink("", params...).
+//
+// See: https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda
+func ResetHyperlink(params ...string) string {
+	return SetHyperlink("", params...)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/iterm2.go b/vendor/github.com/charmbracelet/x/ansi/iterm2.go
new file mode 100644
index 0000000..0ecb336
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/iterm2.go
@@ -0,0 +1,18 @@
+package ansi
+
+import "fmt"
+
+// ITerm2 returns a sequence that uses the iTerm2 proprietary protocol. Use the
+// iterm2 package for a more convenient API.
+//
+//	OSC 1337 ; key = value ST
+//
+// Example:
+//
+//	ITerm2(iterm2.File{...})
+//
+// See https://iterm2.com/documentation-escape-codes.html
+// See https://iterm2.com/documentation-images.html
+func ITerm2(data any) string {
+	return "\x1b]1337;" + fmt.Sprint(data) + "\x07"
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/keypad.go b/vendor/github.com/charmbracelet/x/ansi/keypad.go
new file mode 100644
index 0000000..9183c6a
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/keypad.go
@@ -0,0 +1,28 @@
+package ansi
+
+// Keypad Application Mode (DECKPAM) is a mode that determines whether the
+// keypad sends application sequences or ANSI sequences.
+//
+// This works like enabling [DECNKM].
+// Use [NumericKeypadMode] to set the numeric keypad mode.
+//
+//	ESC =
+//
+// See: https://vt100.net/docs/vt510-rm/DECKPAM.html
+const (
+	KeypadApplicationMode = "\x1b="
+	DECKPAM               = KeypadApplicationMode
+)
+
+// Keypad Numeric Mode (DECKPNM) is a mode that determines whether the keypad
+// sends application sequences or ANSI sequences.
+//
+// This works the same as disabling [DECNKM].
+//
+//	ESC >
+//
+// See: https://vt100.net/docs/vt510-rm/DECKPNM.html
+const (
+	KeypadNumericMode = "\x1b>"
+	DECKPNM           = KeypadNumericMode
+)
diff --git a/vendor/github.com/charmbracelet/x/ansi/kitty.go b/vendor/github.com/charmbracelet/x/ansi/kitty.go
new file mode 100644
index 0000000..124ab83
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/kitty.go
@@ -0,0 +1,90 @@
+package ansi
+
+import "strconv"
+
+// Kitty keyboard protocol progressive enhancement flags.
+// See: https://sw.kovidgoyal.net/kitty/keyboard-protocol/#progressive-enhancement
+const (
+	KittyDisambiguateEscapeCodes = 1 << iota
+	KittyReportEventTypes
+	KittyReportAlternateKeys
+	KittyReportAllKeysAsEscapeCodes
+	KittyReportAssociatedKeys
+
+	KittyAllFlags = KittyDisambiguateEscapeCodes | KittyReportEventTypes |
+		KittyReportAlternateKeys | KittyReportAllKeysAsEscapeCodes | KittyReportAssociatedKeys
+)
+
+// RequestKittyKeyboard is a sequence to request the terminal Kitty keyboard
+// protocol enabled flags.
+//
+// See: https://sw.kovidgoyal.net/kitty/keyboard-protocol/
+const RequestKittyKeyboard = "\x1b[?u"
+
+// KittyKeyboard returns a sequence to request keyboard enhancements from the terminal.
+// The flags argument is a bitmask of the Kitty keyboard protocol flags. While
+// mode specifies how the flags should be interpreted.
+//
+// Possible values for flags mask:
+//
+//	1:  Disambiguate escape codes
+//	2:  Report event types
+//	4:  Report alternate keys
+//	8:  Report all keys as escape codes
+//	16: Report associated text
+//
+// Possible values for mode:
+//
+//	1: Set given flags and unset all others
+//	2: Set given flags and keep existing flags unchanged
+//	3: Unset given flags and keep existing flags unchanged
+//
+// See https://sw.kovidgoyal.net/kitty/keyboard-protocol/#progressive-enhancement
+func KittyKeyboard(flags, mode int) string {
+	return "\x1b[=" + strconv.Itoa(flags) + ";" + strconv.Itoa(mode) + "u"
+}
+
+// PushKittyKeyboard returns a sequence to push the given flags to the terminal
+// Kitty Keyboard stack.
+//
+// Possible values for flags mask:
+//
+//	0:  Disable all features
+//	1:  Disambiguate escape codes
+//	2:  Report event types
+//	4:  Report alternate keys
+//	8:  Report all keys as escape codes
+//	16: Report associated text
+//
+//	CSI > flags u
+//
+// See https://sw.kovidgoyal.net/kitty/keyboard-protocol/#progressive-enhancement
+func PushKittyKeyboard(flags int) string {
+	var f string
+	if flags > 0 {
+		f = strconv.Itoa(flags)
+	}
+
+	return "\x1b[>" + f + "u"
+}
+
+// DisableKittyKeyboard is a sequence to push zero into the terminal Kitty
+// Keyboard stack to disable the protocol.
+//
+// This is equivalent to PushKittyKeyboard(0).
+const DisableKittyKeyboard = "\x1b[>u"
+
+// PopKittyKeyboard returns a sequence to pop n number of flags from the
+// terminal Kitty Keyboard stack.
+//
+//	CSI < flags u
+//
+// See https://sw.kovidgoyal.net/kitty/keyboard-protocol/#progressive-enhancement
+func PopKittyKeyboard(n int) string {
+	var num string
+	if n > 0 {
+		num = strconv.Itoa(n)
+	}
+
+	return "\x1b[<" + num + "u"
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/kitty/decoder.go b/vendor/github.com/charmbracelet/x/ansi/kitty/decoder.go
new file mode 100644
index 0000000..fbd0844
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/kitty/decoder.go
@@ -0,0 +1,85 @@
+package kitty
+
+import (
+	"compress/zlib"
+	"fmt"
+	"image"
+	"image/color"
+	"image/png"
+	"io"
+)
+
+// Decoder is a decoder for the Kitty graphics protocol. It supports decoding
+// images in the 24-bit [RGB], 32-bit [RGBA], and [PNG] formats. It can also
+// decompress data using zlib.
+// The default format is 32-bit [RGBA].
+type Decoder struct {
+	// Uses zlib decompression.
+	Decompress bool
+
+	// Can be one of [RGB], [RGBA], or [PNG].
+	Format int
+
+	// Width of the image in pixels. This can be omitted if the image is [PNG]
+	// formatted.
+	Width int
+
+	// Height of the image in pixels. This can be omitted if the image is [PNG]
+	// formatted.
+	Height int
+}
+
+// Decode decodes the image data from r in the specified format.
+func (d *Decoder) Decode(r io.Reader) (image.Image, error) {
+	if d.Decompress {
+		zr, err := zlib.NewReader(r)
+		if err != nil {
+			return nil, fmt.Errorf("failed to create zlib reader: %w", err)
+		}
+
+		defer zr.Close() //nolint:errcheck
+		r = zr
+	}
+
+	if d.Format == 0 {
+		d.Format = RGBA
+	}
+
+	switch d.Format {
+	case RGBA, RGB:
+		return d.decodeRGBA(r, d.Format == RGBA)
+
+	case PNG:
+		return png.Decode(r)
+
+	default:
+		return nil, fmt.Errorf("unsupported format: %d", d.Format)
+	}
+}
+
+// decodeRGBA decodes the image data in 32-bit RGBA or 24-bit RGB formats.
+func (d *Decoder) decodeRGBA(r io.Reader, alpha bool) (image.Image, error) {
+	m := image.NewRGBA(image.Rect(0, 0, d.Width, d.Height))
+
+	var buf []byte
+	if alpha {
+		buf = make([]byte, 4)
+	} else {
+		buf = make([]byte, 3)
+	}
+
+	for y := 0; y < d.Height; y++ {
+		for x := 0; x < d.Width; x++ {
+			if _, err := io.ReadFull(r, buf[:]); err != nil {
+				return nil, fmt.Errorf("failed to read pixel data: %w", err)
+			}
+			if alpha {
+				m.SetRGBA(x, y, color.RGBA{buf[0], buf[1], buf[2], buf[3]})
+			} else {
+				m.SetRGBA(x, y, color.RGBA{buf[0], buf[1], buf[2], 0xff})
+			}
+		}
+	}
+
+	return m, nil
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/kitty/encoder.go b/vendor/github.com/charmbracelet/x/ansi/kitty/encoder.go
new file mode 100644
index 0000000..f668b9e
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/kitty/encoder.go
@@ -0,0 +1,64 @@
+package kitty
+
+import (
+	"compress/zlib"
+	"fmt"
+	"image"
+	"image/png"
+	"io"
+)
+
+// Encoder is an encoder for the Kitty graphics protocol. It supports encoding
+// images in the 24-bit [RGB], 32-bit [RGBA], and [PNG] formats, and
+// compressing the data using zlib.
+// The default format is 32-bit [RGBA].
+type Encoder struct {
+	// Uses zlib compression.
+	Compress bool
+
+	// Can be one of [RGBA], [RGB], or [PNG].
+	Format int
+}
+
+// Encode encodes the image data in the specified format and writes it to w.
+func (e *Encoder) Encode(w io.Writer, m image.Image) error {
+	if m == nil {
+		return nil
+	}
+
+	if e.Compress {
+		zw := zlib.NewWriter(w)
+		defer zw.Close() //nolint:errcheck
+		w = zw
+	}
+
+	if e.Format == 0 {
+		e.Format = RGBA
+	}
+
+	switch e.Format {
+	case RGBA, RGB:
+		bounds := m.Bounds()
+		for y := bounds.Min.Y; y < bounds.Max.Y; y++ {
+			for x := bounds.Min.X; x < bounds.Max.X; x++ {
+				r, g, b, a := m.At(x, y).RGBA()
+				switch e.Format {
+				case RGBA:
+					w.Write([]byte{byte(r >> 8), byte(g >> 8), byte(b >> 8), byte(a >> 8)}) //nolint:errcheck
+				case RGB:
+					w.Write([]byte{byte(r >> 8), byte(g >> 8), byte(b >> 8)}) //nolint:errcheck
+				}
+			}
+		}
+
+	case PNG:
+		if err := png.Encode(w, m); err != nil {
+			return fmt.Errorf("failed to encode PNG: %w", err)
+		}
+
+	default:
+		return fmt.Errorf("unsupported format: %d", e.Format)
+	}
+
+	return nil
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/kitty/graphics.go b/vendor/github.com/charmbracelet/x/ansi/kitty/graphics.go
new file mode 100644
index 0000000..490e7a8
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/kitty/graphics.go
@@ -0,0 +1,414 @@
+package kitty
+
+import "errors"
+
+// ErrMissingFile is returned when the file path is missing.
+var ErrMissingFile = errors.New("missing file path")
+
+// MaxChunkSize is the maximum chunk size for the image data.
+const MaxChunkSize = 1024 * 4
+
+// Placeholder is a special Unicode character that can be used as a placeholder
+// for an image.
+const Placeholder = '\U0010EEEE'
+
+// Graphics image format.
+const (
+	// 32-bit RGBA format.
+	RGBA = 32
+
+	// 24-bit RGB format.
+	RGB = 24
+
+	// PNG format.
+	PNG = 100
+)
+
+// Compression types.
+const (
+	Zlib = 'z'
+)
+
+// Transmission types.
+const (
+	// The data transmitted directly in the escape sequence.
+	Direct = 'd'
+
+	// The data transmitted in a regular file.
+	File = 'f'
+
+	// A temporary file is used and deleted after transmission.
+	TempFile = 't'
+
+	// A shared memory object.
+	// For POSIX see https://pubs.opengroup.org/onlinepubs/9699919799/functions/shm_open.html
+	// For Windows see https://docs.microsoft.com/en-us/windows/win32/memory/creating-named-shared-memory
+	SharedMemory = 's'
+)
+
+// Action types.
+const (
+	// Transmit image data.
+	Transmit = 't'
+	// TransmitAndPut transmit image data and display (put) it.
+	TransmitAndPut = 'T'
+	// Query terminal for image info.
+	Query = 'q'
+	// Put (display) previously transmitted image.
+	Put = 'p'
+	// Delete image.
+	Delete = 'd'
+	// Frame transmits data for animation frames.
+	Frame = 'f'
+	// Animate controls animation.
+	Animate = 'a'
+	// Compose composes animation frames.
+	Compose = 'c'
+)
+
+// Delete types.
+const (
+	// Delete all placements visible on screen
+	DeleteAll = 'a'
+	// Delete all images with the specified id, specified using the i key. If
+	// you specify a p key for the placement id as well, then only the
+	// placement with the specified image id and placement id will be deleted.
+	DeleteID = 'i'
+	// Delete newest image with the specified number, specified using the I
+	// key. If you specify a p key for the placement id as well, then only the
+	// placement with the specified number and placement id will be deleted.
+	DeleteNumber = 'n'
+	// Delete all placements that intersect with the current cursor position.
+	DeleteCursor = 'c'
+	// Delete animation frames.
+	DeleteFrames = 'f'
+	// Delete all placements that intersect a specific cell, the cell is
+	// specified using the x and y keys
+	DeleteCell = 'p'
+	// Delete all placements that intersect a specific cell having a specific
+	// z-index. The cell and z-index is specified using the x, y and z keys.
+	DeleteCellZ = 'q'
+	// Delete all images whose id is greater than or equal to the value of the x
+	// key and less than or equal to the value of the y.
+	DeleteRange = 'r'
+	// Delete all placements that intersect the specified column, specified using
+	// the x key.
+	DeleteColumn = 'x'
+	// Delete all placements that intersect the specified row, specified using
+	// the y key.
+	DeleteRow = 'y'
+	// Delete all placements that have the specified z-index, specified using the
+	// z key.
+	DeleteZ = 'z'
+)
+
+// Diacritic returns the diacritic rune at the specified index. If the index is
+// out of bounds, the first diacritic rune is returned.
+func Diacritic(i int) rune {
+	if i < 0 || i >= len(diacritics) {
+		return diacritics[0]
+	}
+	return diacritics[i]
+}
+
+// From https://sw.kovidgoyal.net/kitty/_downloads/f0a0de9ec8d9ff4456206db8e0814937/rowcolumn-diacritics.txt
+// See https://sw.kovidgoyal.net/kitty/graphics-protocol/#unicode-placeholders for further explanation.
+var diacritics = []rune{
+	'\u0305',
+	'\u030D',
+	'\u030E',
+	'\u0310',
+	'\u0312',
+	'\u033D',
+	'\u033E',
+	'\u033F',
+	'\u0346',
+	'\u034A',
+	'\u034B',
+	'\u034C',
+	'\u0350',
+	'\u0351',
+	'\u0352',
+	'\u0357',
+	'\u035B',
+	'\u0363',
+	'\u0364',
+	'\u0365',
+	'\u0366',
+	'\u0367',
+	'\u0368',
+	'\u0369',
+	'\u036A',
+	'\u036B',
+	'\u036C',
+	'\u036D',
+	'\u036E',
+	'\u036F',
+	'\u0483',
+	'\u0484',
+	'\u0485',
+	'\u0486',
+	'\u0487',
+	'\u0592',
+	'\u0593',
+	'\u0594',
+	'\u0595',
+	'\u0597',
+	'\u0598',
+	'\u0599',
+	'\u059C',
+	'\u059D',
+	'\u059E',
+	'\u059F',
+	'\u05A0',
+	'\u05A1',
+	'\u05A8',
+	'\u05A9',
+	'\u05AB',
+	'\u05AC',
+	'\u05AF',
+	'\u05C4',
+	'\u0610',
+	'\u0611',
+	'\u0612',
+	'\u0613',
+	'\u0614',
+	'\u0615',
+	'\u0616',
+	'\u0617',
+	'\u0657',
+	'\u0658',
+	'\u0659',
+	'\u065A',
+	'\u065B',
+	'\u065D',
+	'\u065E',
+	'\u06D6',
+	'\u06D7',
+	'\u06D8',
+	'\u06D9',
+	'\u06DA',
+	'\u06DB',
+	'\u06DC',
+	'\u06DF',
+	'\u06E0',
+	'\u06E1',
+	'\u06E2',
+	'\u06E4',
+	'\u06E7',
+	'\u06E8',
+	'\u06EB',
+	'\u06EC',
+	'\u0730',
+	'\u0732',
+	'\u0733',
+	'\u0735',
+	'\u0736',
+	'\u073A',
+	'\u073D',
+	'\u073F',
+	'\u0740',
+	'\u0741',
+	'\u0743',
+	'\u0745',
+	'\u0747',
+	'\u0749',
+	'\u074A',
+	'\u07EB',
+	'\u07EC',
+	'\u07ED',
+	'\u07EE',
+	'\u07EF',
+	'\u07F0',
+	'\u07F1',
+	'\u07F3',
+	'\u0816',
+	'\u0817',
+	'\u0818',
+	'\u0819',
+	'\u081B',
+	'\u081C',
+	'\u081D',
+	'\u081E',
+	'\u081F',
+	'\u0820',
+	'\u0821',
+	'\u0822',
+	'\u0823',
+	'\u0825',
+	'\u0826',
+	'\u0827',
+	'\u0829',
+	'\u082A',
+	'\u082B',
+	'\u082C',
+	'\u082D',
+	'\u0951',
+	'\u0953',
+	'\u0954',
+	'\u0F82',
+	'\u0F83',
+	'\u0F86',
+	'\u0F87',
+	'\u135D',
+	'\u135E',
+	'\u135F',
+	'\u17DD',
+	'\u193A',
+	'\u1A17',
+	'\u1A75',
+	'\u1A76',
+	'\u1A77',
+	'\u1A78',
+	'\u1A79',
+	'\u1A7A',
+	'\u1A7B',
+	'\u1A7C',
+	'\u1B6B',
+	'\u1B6D',
+	'\u1B6E',
+	'\u1B6F',
+	'\u1B70',
+	'\u1B71',
+	'\u1B72',
+	'\u1B73',
+	'\u1CD0',
+	'\u1CD1',
+	'\u1CD2',
+	'\u1CDA',
+	'\u1CDB',
+	'\u1CE0',
+	'\u1DC0',
+	'\u1DC1',
+	'\u1DC3',
+	'\u1DC4',
+	'\u1DC5',
+	'\u1DC6',
+	'\u1DC7',
+	'\u1DC8',
+	'\u1DC9',
+	'\u1DCB',
+	'\u1DCC',
+	'\u1DD1',
+	'\u1DD2',
+	'\u1DD3',
+	'\u1DD4',
+	'\u1DD5',
+	'\u1DD6',
+	'\u1DD7',
+	'\u1DD8',
+	'\u1DD9',
+	'\u1DDA',
+	'\u1DDB',
+	'\u1DDC',
+	'\u1DDD',
+	'\u1DDE',
+	'\u1DDF',
+	'\u1DE0',
+	'\u1DE1',
+	'\u1DE2',
+	'\u1DE3',
+	'\u1DE4',
+	'\u1DE5',
+	'\u1DE6',
+	'\u1DFE',
+	'\u20D0',
+	'\u20D1',
+	'\u20D4',
+	'\u20D5',
+	'\u20D6',
+	'\u20D7',
+	'\u20DB',
+	'\u20DC',
+	'\u20E1',
+	'\u20E7',
+	'\u20E9',
+	'\u20F0',
+	'\u2CEF',
+	'\u2CF0',
+	'\u2CF1',
+	'\u2DE0',
+	'\u2DE1',
+	'\u2DE2',
+	'\u2DE3',
+	'\u2DE4',
+	'\u2DE5',
+	'\u2DE6',
+	'\u2DE7',
+	'\u2DE8',
+	'\u2DE9',
+	'\u2DEA',
+	'\u2DEB',
+	'\u2DEC',
+	'\u2DED',
+	'\u2DEE',
+	'\u2DEF',
+	'\u2DF0',
+	'\u2DF1',
+	'\u2DF2',
+	'\u2DF3',
+	'\u2DF4',
+	'\u2DF5',
+	'\u2DF6',
+	'\u2DF7',
+	'\u2DF8',
+	'\u2DF9',
+	'\u2DFA',
+	'\u2DFB',
+	'\u2DFC',
+	'\u2DFD',
+	'\u2DFE',
+	'\u2DFF',
+	'\uA66F',
+	'\uA67C',
+	'\uA67D',
+	'\uA6F0',
+	'\uA6F1',
+	'\uA8E0',
+	'\uA8E1',
+	'\uA8E2',
+	'\uA8E3',
+	'\uA8E4',
+	'\uA8E5',
+	'\uA8E6',
+	'\uA8E7',
+	'\uA8E8',
+	'\uA8E9',
+	'\uA8EA',
+	'\uA8EB',
+	'\uA8EC',
+	'\uA8ED',
+	'\uA8EE',
+	'\uA8EF',
+	'\uA8F0',
+	'\uA8F1',
+	'\uAAB0',
+	'\uAAB2',
+	'\uAAB3',
+	'\uAAB7',
+	'\uAAB8',
+	'\uAABE',
+	'\uAABF',
+	'\uAAC1',
+	'\uFE20',
+	'\uFE21',
+	'\uFE22',
+	'\uFE23',
+	'\uFE24',
+	'\uFE25',
+	'\uFE26',
+	'\U00010A0F',
+	'\U00010A38',
+	'\U0001D185',
+	'\U0001D186',
+	'\U0001D187',
+	'\U0001D188',
+	'\U0001D189',
+	'\U0001D1AA',
+	'\U0001D1AB',
+	'\U0001D1AC',
+	'\U0001D1AD',
+	'\U0001D242',
+	'\U0001D243',
+	'\U0001D244',
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/kitty/options.go b/vendor/github.com/charmbracelet/x/ansi/kitty/options.go
new file mode 100644
index 0000000..a8d907b
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/kitty/options.go
@@ -0,0 +1,367 @@
+package kitty
+
+import (
+	"encoding"
+	"fmt"
+	"strconv"
+	"strings"
+)
+
+var (
+	_ encoding.TextMarshaler   = Options{}
+	_ encoding.TextUnmarshaler = &Options{}
+)
+
+// Options represents a Kitty Graphics Protocol options.
+type Options struct {
+	// Common options.
+
+	// Action (a=t) is the action to be performed on the image. Can be one of
+	// [Transmit], [TransmitDisplay], [Query], [Put], [Delete], [Frame],
+	// [Animate], [Compose].
+	Action byte
+
+	// Quite mode (q=0) is the quiet mode. Can be either zero, one, or two
+	// where zero is the default, 1 suppresses OK responses, and 2 suppresses
+	// both OK and error responses.
+	Quite byte
+
+	// Transmission options.
+
+	// ID (i=) is the image ID. The ID is a unique identifier for the image.
+	// Must be a positive integer up to [math.MaxUint32].
+	ID int
+
+	// PlacementID (p=) is the placement ID. The placement ID is a unique
+	// identifier for the placement of the image. Must be a positive integer up
+	// to [math.MaxUint32].
+	PlacementID int
+
+	// Number (I=0) is the number of images to be transmitted.
+	Number int
+
+	// Format (f=32) is the image format. One of [RGBA], [RGB], [PNG].
+	Format int
+
+	// ImageWidth (s=0) is the transmitted image width.
+	ImageWidth int
+
+	// ImageHeight (v=0) is the transmitted image height.
+	ImageHeight int
+
+	// Compression (o=) is the image compression type. Can be [Zlib] or zero.
+	Compression byte
+
+	// Transmission (t=d) is the image transmission type. Can be [Direct], [File],
+	// [TempFile], or[SharedMemory].
+	Transmission byte
+
+	// File is the file path to be used when the transmission type is [File].
+	// If [Options.Transmission] is omitted i.e. zero and this is non-empty,
+	// the transmission type is set to [File].
+	File string
+
+	// Size (S=0) is the size to be read from the transmission medium.
+	Size int
+
+	// Offset (O=0) is the offset byte to start reading from the transmission
+	// medium.
+	Offset int
+
+	// Chunk (m=) whether the image is transmitted in chunks. Can be either
+	// zero or one. When true, the image is transmitted in chunks. Each chunk
+	// must be a multiple of 4, and up to [MaxChunkSize] bytes. Each chunk must
+	// have the m=1 option except for the last chunk which must have m=0.
+	Chunk bool
+
+	// Display options.
+
+	// X (x=0) is the pixel X coordinate of the image to start displaying.
+	X int
+
+	// Y (y=0) is the pixel Y coordinate of the image to start displaying.
+	Y int
+
+	// Z (z=0) is the Z coordinate of the image to display.
+	Z int
+
+	// Width (w=0) is the width of the image to display.
+	Width int
+
+	// Height (h=0) is the height of the image to display.
+	Height int
+
+	// OffsetX (X=0) is the OffsetX coordinate of the cursor cell to start
+	// displaying the image. OffsetX=0 is the leftmost cell. This must be
+	// smaller than the terminal cell width.
+	OffsetX int
+
+	// OffsetY (Y=0) is the OffsetY coordinate of the cursor cell to start
+	// displaying the image. OffsetY=0 is the topmost cell. This must be
+	// smaller than the terminal cell height.
+	OffsetY int
+
+	// Columns (c=0) is the number of columns to display the image. The image
+	// will be scaled to fit the number of columns.
+	Columns int
+
+	// Rows (r=0) is the number of rows to display the image. The image will be
+	// scaled to fit the number of rows.
+	Rows int
+
+	// VirtualPlacement (U=0) whether to use virtual placement. This is used
+	// with Unicode [Placeholder] to display images.
+	VirtualPlacement bool
+
+	// DoNotMoveCursor (C=0) whether to move the cursor after displaying the
+	// image.
+	DoNotMoveCursor bool
+
+	// ParentID (P=0) is the parent image ID. The parent ID is the ID of the
+	// image that is the parent of the current image. This is used with Unicode
+	// [Placeholder] to display images relative to the parent image.
+	ParentID int
+
+	// ParentPlacementID (Q=0) is the parent placement ID. The parent placement
+	// ID is the ID of the placement of the parent image. This is used with
+	// Unicode [Placeholder] to display images relative to the parent image.
+	ParentPlacementID int
+
+	// Delete options.
+
+	// Delete (d=a) is the delete action. Can be one of [DeleteAll],
+	// [DeleteID], [DeleteNumber], [DeleteCursor], [DeleteFrames],
+	// [DeleteCell], [DeleteCellZ], [DeleteRange], [DeleteColumn], [DeleteRow],
+	// [DeleteZ].
+	Delete byte
+
+	// DeleteResources indicates whether to delete the resources associated
+	// with the image.
+	DeleteResources bool
+}
+
+// Options returns the options as a slice of a key-value pairs.
+func (o *Options) Options() (opts []string) {
+	opts = []string{}
+	if o.Format == 0 {
+		o.Format = RGBA
+	}
+
+	if o.Action == 0 {
+		o.Action = Transmit
+	}
+
+	if o.Delete == 0 {
+		o.Delete = DeleteAll
+	}
+
+	if o.Transmission == 0 {
+		if len(o.File) > 0 {
+			o.Transmission = File
+		} else {
+			o.Transmission = Direct
+		}
+	}
+
+	if o.Format != RGBA {
+		opts = append(opts, fmt.Sprintf("f=%d", o.Format))
+	}
+
+	if o.Quite > 0 {
+		opts = append(opts, fmt.Sprintf("q=%d", o.Quite))
+	}
+
+	if o.ID > 0 {
+		opts = append(opts, fmt.Sprintf("i=%d", o.ID))
+	}
+
+	if o.PlacementID > 0 {
+		opts = append(opts, fmt.Sprintf("p=%d", o.PlacementID))
+	}
+
+	if o.Number > 0 {
+		opts = append(opts, fmt.Sprintf("I=%d", o.Number))
+	}
+
+	if o.ImageWidth > 0 {
+		opts = append(opts, fmt.Sprintf("s=%d", o.ImageWidth))
+	}
+
+	if o.ImageHeight > 0 {
+		opts = append(opts, fmt.Sprintf("v=%d", o.ImageHeight))
+	}
+
+	if o.Transmission != Direct {
+		opts = append(opts, fmt.Sprintf("t=%c", o.Transmission))
+	}
+
+	if o.Size > 0 {
+		opts = append(opts, fmt.Sprintf("S=%d", o.Size))
+	}
+
+	if o.Offset > 0 {
+		opts = append(opts, fmt.Sprintf("O=%d", o.Offset))
+	}
+
+	if o.Compression == Zlib {
+		opts = append(opts, fmt.Sprintf("o=%c", o.Compression))
+	}
+
+	if o.VirtualPlacement {
+		opts = append(opts, "U=1")
+	}
+
+	if o.DoNotMoveCursor {
+		opts = append(opts, "C=1")
+	}
+
+	if o.ParentID > 0 {
+		opts = append(opts, fmt.Sprintf("P=%d", o.ParentID))
+	}
+
+	if o.ParentPlacementID > 0 {
+		opts = append(opts, fmt.Sprintf("Q=%d", o.ParentPlacementID))
+	}
+
+	if o.X > 0 {
+		opts = append(opts, fmt.Sprintf("x=%d", o.X))
+	}
+
+	if o.Y > 0 {
+		opts = append(opts, fmt.Sprintf("y=%d", o.Y))
+	}
+
+	if o.Z > 0 {
+		opts = append(opts, fmt.Sprintf("z=%d", o.Z))
+	}
+
+	if o.Width > 0 {
+		opts = append(opts, fmt.Sprintf("w=%d", o.Width))
+	}
+
+	if o.Height > 0 {
+		opts = append(opts, fmt.Sprintf("h=%d", o.Height))
+	}
+
+	if o.OffsetX > 0 {
+		opts = append(opts, fmt.Sprintf("X=%d", o.OffsetX))
+	}
+
+	if o.OffsetY > 0 {
+		opts = append(opts, fmt.Sprintf("Y=%d", o.OffsetY))
+	}
+
+	if o.Columns > 0 {
+		opts = append(opts, fmt.Sprintf("c=%d", o.Columns))
+	}
+
+	if o.Rows > 0 {
+		opts = append(opts, fmt.Sprintf("r=%d", o.Rows))
+	}
+
+	if o.Delete != DeleteAll || o.DeleteResources {
+		da := o.Delete
+		if o.DeleteResources {
+			da = da - ' ' // to uppercase
+		}
+
+		opts = append(opts, fmt.Sprintf("d=%c", da))
+	}
+
+	if o.Action != Transmit {
+		opts = append(opts, fmt.Sprintf("a=%c", o.Action))
+	}
+
+	return
+}
+
+// String returns the string representation of the options.
+func (o Options) String() string {
+	return strings.Join(o.Options(), ",")
+}
+
+// MarshalText returns the string representation of the options.
+func (o Options) MarshalText() ([]byte, error) {
+	return []byte(o.String()), nil
+}
+
+// UnmarshalText parses the options from the given string.
+func (o *Options) UnmarshalText(text []byte) error {
+	opts := strings.Split(string(text), ",")
+	for _, opt := range opts {
+		ps := strings.SplitN(opt, "=", 2)
+		if len(ps) != 2 || len(ps[1]) == 0 {
+			continue
+		}
+
+		switch ps[0] {
+		case "a":
+			o.Action = ps[1][0]
+		case "o":
+			o.Compression = ps[1][0]
+		case "t":
+			o.Transmission = ps[1][0]
+		case "d":
+			d := ps[1][0]
+			if d >= 'A' && d <= 'Z' {
+				o.DeleteResources = true
+				d = d + ' ' // to lowercase
+			}
+			o.Delete = d
+		case "i", "q", "p", "I", "f", "s", "v", "S", "O", "m", "x", "y", "z", "w", "h", "X", "Y", "c", "r", "U", "P", "Q":
+			v, err := strconv.Atoi(ps[1])
+			if err != nil {
+				continue
+			}
+
+			switch ps[0] {
+			case "i":
+				o.ID = v
+			case "q":
+				o.Quite = byte(v)
+			case "p":
+				o.PlacementID = v
+			case "I":
+				o.Number = v
+			case "f":
+				o.Format = v
+			case "s":
+				o.ImageWidth = v
+			case "v":
+				o.ImageHeight = v
+			case "S":
+				o.Size = v
+			case "O":
+				o.Offset = v
+			case "m":
+				o.Chunk = v == 0 || v == 1
+			case "x":
+				o.X = v
+			case "y":
+				o.Y = v
+			case "z":
+				o.Z = v
+			case "w":
+				o.Width = v
+			case "h":
+				o.Height = v
+			case "X":
+				o.OffsetX = v
+			case "Y":
+				o.OffsetY = v
+			case "c":
+				o.Columns = v
+			case "r":
+				o.Rows = v
+			case "U":
+				o.VirtualPlacement = v == 1
+			case "P":
+				o.ParentID = v
+			case "Q":
+				o.ParentPlacementID = v
+			}
+		}
+	}
+
+	return nil
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/method.go b/vendor/github.com/charmbracelet/x/ansi/method.go
new file mode 100644
index 0000000..0218809
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/method.go
@@ -0,0 +1,172 @@
+package ansi
+
+// Method is a type that represents the how the renderer should calculate the
+// display width of cells.
+type Method uint8
+
+// Display width modes.
+const (
+	WcWidth Method = iota
+	GraphemeWidth
+)
+
+// StringWidth returns the width of a string in cells. This is the number of
+// cells that the string will occupy when printed in a terminal. ANSI escape
+// codes are ignored and wide characters (such as East Asians and emojis) are
+// accounted for.
+func (m Method) StringWidth(s string) int {
+	return stringWidth(m, s)
+}
+
+// Truncate truncates a string to a given length, adding a tail to the end if
+// the string is longer than the given length. This function is aware of ANSI
+// escape codes and will not break them, and accounts for wide-characters (such
+// as East-Asian characters and emojis).
+func (m Method) Truncate(s string, length int, tail string) string {
+	return truncate(m, s, length, tail)
+}
+
+// TruncateLeft truncates a string to a given length, adding a prefix to the
+// beginning if the string is longer than the given length. This function is
+// aware of ANSI escape codes and will not break them, and accounts for
+// wide-characters (such as East-Asian characters and emojis).
+func (m Method) TruncateLeft(s string, length int, prefix string) string {
+	return truncateLeft(m, s, length, prefix)
+}
+
+// Cut the string, without adding any prefix or tail strings. This function is
+// aware of ANSI escape codes and will not break them, and accounts for
+// wide-characters (such as East-Asian characters and emojis). Note that the
+// [left] parameter is inclusive, while [right] isn't.
+func (m Method) Cut(s string, left, right int) string {
+	return cut(m, s, left, right)
+}
+
+// Hardwrap wraps a string or a block of text to a given line length, breaking
+// word boundaries. This will preserve ANSI escape codes and will account for
+// wide-characters in the string.
+// When preserveSpace is true, spaces at the beginning of a line will be
+// preserved.
+// This treats the text as a sequence of graphemes.
+func (m Method) Hardwrap(s string, length int, preserveSpace bool) string {
+	return hardwrap(m, s, length, preserveSpace)
+}
+
+// Wordwrap wraps a string or a block of text to a given line length, not
+// breaking word boundaries. This will preserve ANSI escape codes and will
+// account for wide-characters in the string.
+// The breakpoints string is a list of characters that are considered
+// breakpoints for word wrapping. A hyphen (-) is always considered a
+// breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+func (m Method) Wordwrap(s string, length int, breakpoints string) string {
+	return wordwrap(m, s, length, breakpoints)
+}
+
+// Wrap wraps a string or a block of text to a given line length, breaking word
+// boundaries if necessary. This will preserve ANSI escape codes and will
+// account for wide-characters in the string. The breakpoints string is a list
+// of characters that are considered breakpoints for word wrapping. A hyphen
+// (-) is always considered a breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+func (m Method) Wrap(s string, length int, breakpoints string) string {
+	return wrap(m, s, length, breakpoints)
+}
+
+// DecodeSequence decodes the first ANSI escape sequence or a printable
+// grapheme from the given data. It returns the sequence slice, the number of
+// bytes read, the cell width for each sequence, and the new state.
+//
+// The cell width will always be 0 for control and escape sequences, 1 for
+// ASCII printable characters, and the number of cells other Unicode characters
+// occupy. It uses the uniseg package to calculate the width of Unicode
+// graphemes and characters. This means it will always do grapheme clustering
+// (mode 2027).
+//
+// Passing a non-nil [*Parser] as the last argument will allow the decoder to
+// collect sequence parameters, data, and commands. The parser cmd will have
+// the packed command value that contains intermediate and prefix characters.
+// In the case of a OSC sequence, the cmd will be the OSC command number. Use
+// [Cmd] and [Param] types to unpack command intermediates and prefixes as well
+// as parameters.
+//
+// Zero [Cmd] means the CSI, DCS, or ESC sequence is invalid. Moreover, checking the
+// validity of other data sequences, OSC, DCS, etc, will require checking for
+// the returned sequence terminator bytes such as ST (ESC \\) and BEL).
+//
+// We store the command byte in [Cmd] in the most significant byte, the
+// prefix byte in the next byte, and the intermediate byte in the least
+// significant byte. This is done to avoid using a struct to store the command
+// and its intermediates and prefixes. The command byte is always the least
+// significant byte i.e. [Cmd & 0xff]. Use the [Cmd] type to unpack the
+// command, intermediate, and prefix bytes. Note that we only collect the last
+// prefix character and intermediate byte.
+//
+// The [p.Params] slice will contain the parameters of the sequence. Any
+// sub-parameter will have the [parser.HasMoreFlag] set. Use the [Param] type
+// to unpack the parameters.
+//
+// Example:
+//
+//	var state byte // the initial state is always zero [NormalState]
+//	p := NewParser(32, 1024) // create a new parser with a 32 params buffer and 1024 data buffer (optional)
+//	input := []byte("\x1b[31mHello, World!\x1b[0m")
+//	for len(input) > 0 {
+//		seq, width, n, newState := DecodeSequence(input, state, p)
+//		log.Printf("seq: %q, width: %d", seq, width)
+//		state = newState
+//		input = input[n:]
+//	}
+func (m Method) DecodeSequence(data []byte, state byte, p *Parser) (seq []byte, width, n int, newState byte) {
+	return decodeSequence(m, data, state, p)
+}
+
+// DecodeSequenceInString decodes the first ANSI escape sequence or a printable
+// grapheme from the given data. It returns the sequence slice, the number of
+// bytes read, the cell width for each sequence, and the new state.
+//
+// The cell width will always be 0 for control and escape sequences, 1 for
+// ASCII printable characters, and the number of cells other Unicode characters
+// occupy. It uses the uniseg package to calculate the width of Unicode
+// graphemes and characters. This means it will always do grapheme clustering
+// (mode 2027).
+//
+// Passing a non-nil [*Parser] as the last argument will allow the decoder to
+// collect sequence parameters, data, and commands. The parser cmd will have
+// the packed command value that contains intermediate and prefix characters.
+// In the case of a OSC sequence, the cmd will be the OSC command number. Use
+// [Cmd] and [Param] types to unpack command intermediates and prefixes as well
+// as parameters.
+//
+// Zero [Cmd] means the CSI, DCS, or ESC sequence is invalid. Moreover, checking the
+// validity of other data sequences, OSC, DCS, etc, will require checking for
+// the returned sequence terminator bytes such as ST (ESC \\) and BEL).
+//
+// We store the command byte in [Cmd] in the most significant byte, the
+// prefix byte in the next byte, and the intermediate byte in the least
+// significant byte. This is done to avoid using a struct to store the command
+// and its intermediates and prefixes. The command byte is always the least
+// significant byte i.e. [Cmd & 0xff]. Use the [Cmd] type to unpack the
+// command, intermediate, and prefix bytes. Note that we only collect the last
+// prefix character and intermediate byte.
+//
+// The [p.Params] slice will contain the parameters of the sequence. Any
+// sub-parameter will have the [parser.HasMoreFlag] set. Use the [Param] type
+// to unpack the parameters.
+//
+// Example:
+//
+//	var state byte // the initial state is always zero [NormalState]
+//	p := NewParser(32, 1024) // create a new parser with a 32 params buffer and 1024 data buffer (optional)
+//	input := []byte("\x1b[31mHello, World!\x1b[0m")
+//	for len(input) > 0 {
+//		seq, width, n, newState := DecodeSequenceInString(input, state, p)
+//		log.Printf("seq: %q, width: %d", seq, width)
+//		state = newState
+//		input = input[n:]
+//	}
+func (m Method) DecodeSequenceInString(data string, state byte, p *Parser) (seq string, width, n int, newState byte) {
+	return decodeSequence(m, data, state, p)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/mode.go b/vendor/github.com/charmbracelet/x/ansi/mode.go
new file mode 100644
index 0000000..57f3f0a
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/mode.go
@@ -0,0 +1,763 @@
+package ansi
+
+import (
+	"strconv"
+	"strings"
+)
+
+// ModeSetting represents a mode setting.
+type ModeSetting byte
+
+// ModeSetting constants.
+const (
+	ModeNotRecognized ModeSetting = iota
+	ModeSet
+	ModeReset
+	ModePermanentlySet
+	ModePermanentlyReset
+)
+
+// IsNotRecognized returns true if the mode is not recognized.
+func (m ModeSetting) IsNotRecognized() bool {
+	return m == ModeNotRecognized
+}
+
+// IsSet returns true if the mode is set or permanently set.
+func (m ModeSetting) IsSet() bool {
+	return m == ModeSet || m == ModePermanentlySet
+}
+
+// IsReset returns true if the mode is reset or permanently reset.
+func (m ModeSetting) IsReset() bool {
+	return m == ModeReset || m == ModePermanentlyReset
+}
+
+// IsPermanentlySet returns true if the mode is permanently set.
+func (m ModeSetting) IsPermanentlySet() bool {
+	return m == ModePermanentlySet
+}
+
+// IsPermanentlyReset returns true if the mode is permanently reset.
+func (m ModeSetting) IsPermanentlyReset() bool {
+	return m == ModePermanentlyReset
+}
+
+// Mode represents an interface for terminal modes.
+// Modes can be set, reset, and requested.
+type Mode interface {
+	Mode() int
+}
+
+// SetMode (SM) returns a sequence to set a mode.
+// The mode arguments are a list of modes to set.
+//
+// If one of the modes is a [DECMode], the function will returns two escape
+// sequences.
+//
+// ANSI format:
+//
+//	CSI Pd ; ... ; Pd h
+//
+// DEC format:
+//
+//	CSI ? Pd ; ... ; Pd h
+//
+// See: https://vt100.net/docs/vt510-rm/SM.html
+func SetMode(modes ...Mode) string {
+	return setMode(false, modes...)
+}
+
+// SM is an alias for [SetMode].
+func SM(modes ...Mode) string {
+	return SetMode(modes...)
+}
+
+// ResetMode (RM) returns a sequence to reset a mode.
+// The mode arguments are a list of modes to reset.
+//
+// If one of the modes is a [DECMode], the function will returns two escape
+// sequences.
+//
+// ANSI format:
+//
+//	CSI Pd ; ... ; Pd l
+//
+// DEC format:
+//
+//	CSI ? Pd ; ... ; Pd l
+//
+// See: https://vt100.net/docs/vt510-rm/RM.html
+func ResetMode(modes ...Mode) string {
+	return setMode(true, modes...)
+}
+
+// RM is an alias for [ResetMode].
+func RM(modes ...Mode) string {
+	return ResetMode(modes...)
+}
+
+func setMode(reset bool, modes ...Mode) (s string) {
+	if len(modes) == 0 {
+		return
+	}
+
+	cmd := "h"
+	if reset {
+		cmd = "l"
+	}
+
+	seq := "\x1b["
+	if len(modes) == 1 {
+		switch modes[0].(type) {
+		case DECMode:
+			seq += "?"
+		}
+		return seq + strconv.Itoa(modes[0].Mode()) + cmd
+	}
+
+	dec := make([]string, 0, len(modes)/2)
+	ansi := make([]string, 0, len(modes)/2)
+	for _, m := range modes {
+		switch m.(type) {
+		case DECMode:
+			dec = append(dec, strconv.Itoa(m.Mode()))
+		case ANSIMode:
+			ansi = append(ansi, strconv.Itoa(m.Mode()))
+		}
+	}
+
+	if len(ansi) > 0 {
+		s += seq + strings.Join(ansi, ";") + cmd
+	}
+	if len(dec) > 0 {
+		s += seq + "?" + strings.Join(dec, ";") + cmd
+	}
+	return
+}
+
+// RequestMode (DECRQM) returns a sequence to request a mode from the terminal.
+// The terminal responds with a report mode function [DECRPM].
+//
+// ANSI format:
+//
+//	CSI Pa $ p
+//
+// DEC format:
+//
+//	CSI ? Pa $ p
+//
+// See: https://vt100.net/docs/vt510-rm/DECRQM.html
+func RequestMode(m Mode) string {
+	seq := "\x1b["
+	switch m.(type) {
+	case DECMode:
+		seq += "?"
+	}
+	return seq + strconv.Itoa(m.Mode()) + "$p"
+}
+
+// DECRQM is an alias for [RequestMode].
+func DECRQM(m Mode) string {
+	return RequestMode(m)
+}
+
+// ReportMode (DECRPM) returns a sequence that the terminal sends to the host
+// in response to a mode request [DECRQM].
+//
+// ANSI format:
+//
+//	CSI Pa ; Ps ; $ y
+//
+// DEC format:
+//
+//	CSI ? Pa ; Ps $ y
+//
+// Where Pa is the mode number, and Ps is the mode value.
+//
+//	0: Not recognized
+//	1: Set
+//	2: Reset
+//	3: Permanent set
+//	4: Permanent reset
+//
+// See: https://vt100.net/docs/vt510-rm/DECRPM.html
+func ReportMode(mode Mode, value ModeSetting) string {
+	if value > 4 {
+		value = 0
+	}
+	switch mode.(type) {
+	case DECMode:
+		return "\x1b[?" + strconv.Itoa(mode.Mode()) + ";" + strconv.Itoa(int(value)) + "$y"
+	}
+	return "\x1b[" + strconv.Itoa(mode.Mode()) + ";" + strconv.Itoa(int(value)) + "$y"
+}
+
+// DECRPM is an alias for [ReportMode].
+func DECRPM(mode Mode, value ModeSetting) string {
+	return ReportMode(mode, value)
+}
+
+// ANSIMode represents an ANSI terminal mode.
+type ANSIMode int //nolint:revive
+
+// Mode returns the ANSI mode as an integer.
+func (m ANSIMode) Mode() int {
+	return int(m)
+}
+
+// DECMode represents a private DEC terminal mode.
+type DECMode int
+
+// Mode returns the DEC mode as an integer.
+func (m DECMode) Mode() int {
+	return int(m)
+}
+
+// Keyboard Action Mode (KAM) is a mode that controls locking of the keyboard.
+// When the keyboard is locked, it cannot send data to the terminal.
+//
+// See: https://vt100.net/docs/vt510-rm/KAM.html
+const (
+	KeyboardActionMode = ANSIMode(2)
+	KAM                = KeyboardActionMode
+
+	SetKeyboardActionMode     = "\x1b[2h"
+	ResetKeyboardActionMode   = "\x1b[2l"
+	RequestKeyboardActionMode = "\x1b[2$p"
+)
+
+// Insert/Replace Mode (IRM) is a mode that determines whether characters are
+// inserted or replaced when typed.
+//
+// When enabled, characters are inserted at the cursor position pushing the
+// characters to the right. When disabled, characters replace the character at
+// the cursor position.
+//
+// See: https://vt100.net/docs/vt510-rm/IRM.html
+const (
+	InsertReplaceMode = ANSIMode(4)
+	IRM               = InsertReplaceMode
+
+	SetInsertReplaceMode     = "\x1b[4h"
+	ResetInsertReplaceMode   = "\x1b[4l"
+	RequestInsertReplaceMode = "\x1b[4$p"
+)
+
+// Send Receive Mode (SRM) or Local Echo Mode is a mode that determines whether
+// the terminal echoes characters back to the host. When enabled, the terminal
+// sends characters to the host as they are typed.
+//
+// See: https://vt100.net/docs/vt510-rm/SRM.html
+const (
+	SendReceiveMode = ANSIMode(12)
+	LocalEchoMode   = SendReceiveMode
+	SRM             = SendReceiveMode
+
+	SetSendReceiveMode     = "\x1b[12h"
+	ResetSendReceiveMode   = "\x1b[12l"
+	RequestSendReceiveMode = "\x1b[12$p"
+
+	SetLocalEchoMode     = "\x1b[12h"
+	ResetLocalEchoMode   = "\x1b[12l"
+	RequestLocalEchoMode = "\x1b[12$p"
+)
+
+// Line Feed/New Line Mode (LNM) is a mode that determines whether the terminal
+// interprets the line feed character as a new line.
+//
+// When enabled, the terminal interprets the line feed character as a new line.
+// When disabled, the terminal interprets the line feed character as a line feed.
+//
+// A new line moves the cursor to the first position of the next line.
+// A line feed moves the cursor down one line without changing the column
+// scrolling the screen if necessary.
+//
+// See: https://vt100.net/docs/vt510-rm/LNM.html
+const (
+	LineFeedNewLineMode = ANSIMode(20)
+	LNM                 = LineFeedNewLineMode
+
+	SetLineFeedNewLineMode     = "\x1b[20h"
+	ResetLineFeedNewLineMode   = "\x1b[20l"
+	RequestLineFeedNewLineMode = "\x1b[20$p"
+)
+
+// Cursor Keys Mode (DECCKM) is a mode that determines whether the cursor keys
+// send ANSI cursor sequences or application sequences.
+//
+// See: https://vt100.net/docs/vt510-rm/DECCKM.html
+const (
+	CursorKeysMode = DECMode(1)
+	DECCKM         = CursorKeysMode
+
+	SetCursorKeysMode     = "\x1b[?1h"
+	ResetCursorKeysMode   = "\x1b[?1l"
+	RequestCursorKeysMode = "\x1b[?1$p"
+)
+
+// Deprecated: use [SetCursorKeysMode] and [ResetCursorKeysMode] instead.
+const (
+	EnableCursorKeys  = "\x1b[?1h"
+	DisableCursorKeys = "\x1b[?1l"
+)
+
+// Origin Mode (DECOM) is a mode that determines whether the cursor moves to the
+// home position or the margin position.
+//
+// See: https://vt100.net/docs/vt510-rm/DECOM.html
+const (
+	OriginMode = DECMode(6)
+	DECOM      = OriginMode
+
+	SetOriginMode     = "\x1b[?6h"
+	ResetOriginMode   = "\x1b[?6l"
+	RequestOriginMode = "\x1b[?6$p"
+)
+
+// Auto Wrap Mode (DECAWM) is a mode that determines whether the cursor wraps
+// to the next line when it reaches the right margin.
+//
+// See: https://vt100.net/docs/vt510-rm/DECAWM.html
+const (
+	AutoWrapMode = DECMode(7)
+	DECAWM       = AutoWrapMode
+
+	SetAutoWrapMode     = "\x1b[?7h"
+	ResetAutoWrapMode   = "\x1b[?7l"
+	RequestAutoWrapMode = "\x1b[?7$p"
+)
+
+// X10 Mouse Mode is a mode that determines whether the mouse reports on button
+// presses.
+//
+// The terminal responds with the following encoding:
+//
+//	CSI M CbCxCy
+//
+// Where Cb is the button-1, where it can be 1, 2, or 3.
+// Cx and Cy are the x and y coordinates of the mouse event.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	X10MouseMode = DECMode(9)
+
+	SetX10MouseMode     = "\x1b[?9h"
+	ResetX10MouseMode   = "\x1b[?9l"
+	RequestX10MouseMode = "\x1b[?9$p"
+)
+
+// Text Cursor Enable Mode (DECTCEM) is a mode that shows/hides the cursor.
+//
+// See: https://vt100.net/docs/vt510-rm/DECTCEM.html
+const (
+	TextCursorEnableMode = DECMode(25)
+	DECTCEM              = TextCursorEnableMode
+
+	SetTextCursorEnableMode     = "\x1b[?25h"
+	ResetTextCursorEnableMode   = "\x1b[?25l"
+	RequestTextCursorEnableMode = "\x1b[?25$p"
+)
+
+// These are aliases for [SetTextCursorEnableMode] and [ResetTextCursorEnableMode].
+const (
+	ShowCursor = SetTextCursorEnableMode
+	HideCursor = ResetTextCursorEnableMode
+)
+
+// Text Cursor Enable Mode (DECTCEM) is a mode that shows/hides the cursor.
+//
+// See: https://vt100.net/docs/vt510-rm/DECTCEM.html
+//
+// Deprecated: use [SetTextCursorEnableMode] and [ResetTextCursorEnableMode] instead.
+const (
+	CursorEnableMode        = DECMode(25)
+	RequestCursorVisibility = "\x1b[?25$p"
+)
+
+// Numeric Keypad Mode (DECNKM) is a mode that determines whether the keypad
+// sends application sequences or numeric sequences.
+//
+// This works like [DECKPAM] and [DECKPNM], but uses different sequences.
+//
+// See: https://vt100.net/docs/vt510-rm/DECNKM.html
+const (
+	NumericKeypadMode = DECMode(66)
+	DECNKM            = NumericKeypadMode
+
+	SetNumericKeypadMode     = "\x1b[?66h"
+	ResetNumericKeypadMode   = "\x1b[?66l"
+	RequestNumericKeypadMode = "\x1b[?66$p"
+)
+
+// Backarrow Key Mode (DECBKM) is a mode that determines whether the backspace
+// key sends a backspace or delete character. Disabled by default.
+//
+// See: https://vt100.net/docs/vt510-rm/DECBKM.html
+const (
+	BackarrowKeyMode = DECMode(67)
+	DECBKM           = BackarrowKeyMode
+
+	SetBackarrowKeyMode     = "\x1b[?67h"
+	ResetBackarrowKeyMode   = "\x1b[?67l"
+	RequestBackarrowKeyMode = "\x1b[?67$p"
+)
+
+// Left Right Margin Mode (DECLRMM) is a mode that determines whether the left
+// and right margins can be set with [DECSLRM].
+//
+// See: https://vt100.net/docs/vt510-rm/DECLRMM.html
+const (
+	LeftRightMarginMode = DECMode(69)
+	DECLRMM             = LeftRightMarginMode
+
+	SetLeftRightMarginMode     = "\x1b[?69h"
+	ResetLeftRightMarginMode   = "\x1b[?69l"
+	RequestLeftRightMarginMode = "\x1b[?69$p"
+)
+
+// Normal Mouse Mode is a mode that determines whether the mouse reports on
+// button presses and releases. It will also report modifier keys, wheel
+// events, and extra buttons.
+//
+// It uses the same encoding as [X10MouseMode] with a few differences:
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	NormalMouseMode = DECMode(1000)
+
+	SetNormalMouseMode     = "\x1b[?1000h"
+	ResetNormalMouseMode   = "\x1b[?1000l"
+	RequestNormalMouseMode = "\x1b[?1000$p"
+)
+
+// VT Mouse Tracking is a mode that determines whether the mouse reports on
+// button press and release.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+//
+// Deprecated: use [NormalMouseMode] instead.
+const (
+	MouseMode = DECMode(1000)
+
+	EnableMouse  = "\x1b[?1000h"
+	DisableMouse = "\x1b[?1000l"
+	RequestMouse = "\x1b[?1000$p"
+)
+
+// Highlight Mouse Tracking is a mode that determines whether the mouse reports
+// on button presses, releases, and highlighted cells.
+//
+// It uses the same encoding as [NormalMouseMode] with a few differences:
+//
+// On highlight events, the terminal responds with the following encoding:
+//
+//	CSI t CxCy
+//	CSI T CxCyCxCyCxCy
+//
+// Where the parameters are startx, starty, endx, endy, mousex, and mousey.
+const (
+	HighlightMouseMode = DECMode(1001)
+
+	SetHighlightMouseMode     = "\x1b[?1001h"
+	ResetHighlightMouseMode   = "\x1b[?1001l"
+	RequestHighlightMouseMode = "\x1b[?1001$p"
+)
+
+// VT Hilite Mouse Tracking is a mode that determines whether the mouse reports on
+// button presses, releases, and highlighted cells.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+//
+// Deprecated: use [HighlightMouseMode] instead.
+const (
+	MouseHiliteMode = DECMode(1001)
+
+	EnableMouseHilite  = "\x1b[?1001h"
+	DisableMouseHilite = "\x1b[?1001l"
+	RequestMouseHilite = "\x1b[?1001$p"
+)
+
+// Button Event Mouse Tracking is essentially the same as [NormalMouseMode],
+// but it also reports button-motion events when a button is pressed.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	ButtonEventMouseMode = DECMode(1002)
+
+	SetButtonEventMouseMode     = "\x1b[?1002h"
+	ResetButtonEventMouseMode   = "\x1b[?1002l"
+	RequestButtonEventMouseMode = "\x1b[?1002$p"
+)
+
+// Cell Motion Mouse Tracking is a mode that determines whether the mouse
+// reports on button press, release, and motion events.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+//
+// Deprecated: use [ButtonEventMouseMode] instead.
+const (
+	MouseCellMotionMode = DECMode(1002)
+
+	EnableMouseCellMotion  = "\x1b[?1002h"
+	DisableMouseCellMotion = "\x1b[?1002l"
+	RequestMouseCellMotion = "\x1b[?1002$p"
+)
+
+// Any Event Mouse Tracking is the same as [ButtonEventMouseMode], except that
+// all motion events are reported even if no mouse buttons are pressed.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	AnyEventMouseMode = DECMode(1003)
+
+	SetAnyEventMouseMode     = "\x1b[?1003h"
+	ResetAnyEventMouseMode   = "\x1b[?1003l"
+	RequestAnyEventMouseMode = "\x1b[?1003$p"
+)
+
+// All Mouse Tracking is a mode that determines whether the mouse reports on
+// button press, release, motion, and highlight events.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+//
+// Deprecated: use [AnyEventMouseMode] instead.
+const (
+	MouseAllMotionMode = DECMode(1003)
+
+	EnableMouseAllMotion  = "\x1b[?1003h"
+	DisableMouseAllMotion = "\x1b[?1003l"
+	RequestMouseAllMotion = "\x1b[?1003$p"
+)
+
+// Focus Event Mode is a mode that determines whether the terminal reports focus
+// and blur events.
+//
+// The terminal sends the following encoding:
+//
+//	CSI I // Focus In
+//	CSI O // Focus Out
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Focus-Tracking
+const (
+	FocusEventMode = DECMode(1004)
+
+	SetFocusEventMode     = "\x1b[?1004h"
+	ResetFocusEventMode   = "\x1b[?1004l"
+	RequestFocusEventMode = "\x1b[?1004$p"
+)
+
+// Deprecated: use [SetFocusEventMode], [ResetFocusEventMode], and
+// [RequestFocusEventMode] instead.
+const (
+	ReportFocusMode = DECMode(1004)
+
+	EnableReportFocus  = "\x1b[?1004h"
+	DisableReportFocus = "\x1b[?1004l"
+	RequestReportFocus = "\x1b[?1004$p"
+)
+
+// SGR Extended Mouse Mode is a mode that changes the mouse tracking encoding
+// to use SGR parameters.
+//
+// The terminal responds with the following encoding:
+//
+//	CSI < Cb ; Cx ; Cy M
+//
+// Where Cb is the same as [NormalMouseMode], and Cx and Cy are the x and y.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	SgrExtMouseMode = DECMode(1006)
+
+	SetSgrExtMouseMode     = "\x1b[?1006h"
+	ResetSgrExtMouseMode   = "\x1b[?1006l"
+	RequestSgrExtMouseMode = "\x1b[?1006$p"
+)
+
+// Deprecated: use [SgrExtMouseMode] [SetSgrExtMouseMode],
+// [ResetSgrExtMouseMode], and [RequestSgrExtMouseMode] instead.
+const (
+	MouseSgrExtMode    = DECMode(1006)
+	EnableMouseSgrExt  = "\x1b[?1006h"
+	DisableMouseSgrExt = "\x1b[?1006l"
+	RequestMouseSgrExt = "\x1b[?1006$p"
+)
+
+// UTF-8 Extended Mouse Mode is a mode that changes the mouse tracking encoding
+// to use UTF-8 parameters.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	Utf8ExtMouseMode = DECMode(1005)
+
+	SetUtf8ExtMouseMode     = "\x1b[?1005h"
+	ResetUtf8ExtMouseMode   = "\x1b[?1005l"
+	RequestUtf8ExtMouseMode = "\x1b[?1005$p"
+)
+
+// URXVT Extended Mouse Mode is a mode that changes the mouse tracking encoding
+// to use an alternate encoding.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	UrxvtExtMouseMode = DECMode(1015)
+
+	SetUrxvtExtMouseMode     = "\x1b[?1015h"
+	ResetUrxvtExtMouseMode   = "\x1b[?1015l"
+	RequestUrxvtExtMouseMode = "\x1b[?1015$p"
+)
+
+// SGR Pixel Extended Mouse Mode is a mode that changes the mouse tracking
+// encoding to use SGR parameters with pixel coordinates.
+//
+// This is similar to [SgrExtMouseMode], but also reports pixel coordinates.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Mouse-Tracking
+const (
+	SgrPixelExtMouseMode = DECMode(1016)
+
+	SetSgrPixelExtMouseMode     = "\x1b[?1016h"
+	ResetSgrPixelExtMouseMode   = "\x1b[?1016l"
+	RequestSgrPixelExtMouseMode = "\x1b[?1016$p"
+)
+
+// Alternate Screen Mode is a mode that determines whether the alternate screen
+// buffer is active. When this mode is enabled, the alternate screen buffer is
+// cleared.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-The-Alternate-Screen-Buffer
+const (
+	AltScreenMode = DECMode(1047)
+
+	SetAltScreenMode     = "\x1b[?1047h"
+	ResetAltScreenMode   = "\x1b[?1047l"
+	RequestAltScreenMode = "\x1b[?1047$p"
+)
+
+// Save Cursor Mode is a mode that saves the cursor position.
+// This is equivalent to [SaveCursor] and [RestoreCursor].
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-The-Alternate-Screen-Buffer
+const (
+	SaveCursorMode = DECMode(1048)
+
+	SetSaveCursorMode     = "\x1b[?1048h"
+	ResetSaveCursorMode   = "\x1b[?1048l"
+	RequestSaveCursorMode = "\x1b[?1048$p"
+)
+
+// Alternate Screen Save Cursor Mode is a mode that saves the cursor position as in
+// [SaveCursorMode], switches to the alternate screen buffer as in [AltScreenMode],
+// and clears the screen on switch.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-The-Alternate-Screen-Buffer
+const (
+	AltScreenSaveCursorMode = DECMode(1049)
+
+	SetAltScreenSaveCursorMode     = "\x1b[?1049h"
+	ResetAltScreenSaveCursorMode   = "\x1b[?1049l"
+	RequestAltScreenSaveCursorMode = "\x1b[?1049$p"
+)
+
+// Alternate Screen Buffer is a mode that determines whether the alternate screen
+// buffer is active.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-The-Alternate-Screen-Buffer
+//
+// Deprecated: use [AltScreenSaveCursorMode] instead.
+const (
+	AltScreenBufferMode = DECMode(1049)
+
+	SetAltScreenBufferMode     = "\x1b[?1049h"
+	ResetAltScreenBufferMode   = "\x1b[?1049l"
+	RequestAltScreenBufferMode = "\x1b[?1049$p"
+
+	EnableAltScreenBuffer  = "\x1b[?1049h"
+	DisableAltScreenBuffer = "\x1b[?1049l"
+	RequestAltScreenBuffer = "\x1b[?1049$p"
+)
+
+// Bracketed Paste Mode is a mode that determines whether pasted text is
+// bracketed with escape sequences.
+//
+// See: https://cirw.in/blog/bracketed-paste
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Bracketed-Paste-Mode
+const (
+	BracketedPasteMode = DECMode(2004)
+
+	SetBracketedPasteMode     = "\x1b[?2004h"
+	ResetBracketedPasteMode   = "\x1b[?2004l"
+	RequestBracketedPasteMode = "\x1b[?2004$p"
+)
+
+// Deprecated: use [SetBracketedPasteMode], [ResetBracketedPasteMode], and
+// [RequestBracketedPasteMode] instead.
+const (
+	EnableBracketedPaste  = "\x1b[?2004h"
+	DisableBracketedPaste = "\x1b[?2004l"
+	RequestBracketedPaste = "\x1b[?2004$p"
+)
+
+// Synchronized Output Mode is a mode that determines whether output is
+// synchronized with the terminal.
+//
+// See: https://gist.github.com/christianparpart/d8a62cc1ab659194337d73e399004036
+const (
+	SynchronizedOutputMode = DECMode(2026)
+
+	SetSynchronizedOutputMode     = "\x1b[?2026h"
+	ResetSynchronizedOutputMode   = "\x1b[?2026l"
+	RequestSynchronizedOutputMode = "\x1b[?2026$p"
+)
+
+// Deprecated: use [SynchronizedOutputMode], [SetSynchronizedOutputMode], and
+// [ResetSynchronizedOutputMode], and [RequestSynchronizedOutputMode] instead.
+const (
+	SyncdOutputMode = DECMode(2026)
+
+	EnableSyncdOutput  = "\x1b[?2026h"
+	DisableSyncdOutput = "\x1b[?2026l"
+	RequestSyncdOutput = "\x1b[?2026$p"
+)
+
+// Grapheme Clustering Mode is a mode that determines whether the terminal
+// should look for grapheme clusters instead of single runes in the rendered
+// text. This makes the terminal properly render combining characters such as
+// emojis.
+//
+// See: https://github.com/contour-terminal/terminal-unicode-core
+const (
+	GraphemeClusteringMode = DECMode(2027)
+
+	SetGraphemeClusteringMode     = "\x1b[?2027h"
+	ResetGraphemeClusteringMode   = "\x1b[?2027l"
+	RequestGraphemeClusteringMode = "\x1b[?2027$p"
+)
+
+// Deprecated: use [SetGraphemeClusteringMode], [ResetGraphemeClusteringMode], and
+// [RequestGraphemeClusteringMode] instead.
+const (
+	EnableGraphemeClustering  = "\x1b[?2027h"
+	DisableGraphemeClustering = "\x1b[?2027l"
+	RequestGraphemeClustering = "\x1b[?2027$p"
+)
+
+// Win32Input is a mode that determines whether input is processed by the
+// Win32 console and Conpty.
+//
+// See: https://github.com/microsoft/terminal/blob/main/doc/specs/%234999%20-%20Improved%20keyboard%20handling%20in%20Conpty.md
+const (
+	Win32InputMode = DECMode(9001)
+
+	SetWin32InputMode     = "\x1b[?9001h"
+	ResetWin32InputMode   = "\x1b[?9001l"
+	RequestWin32InputMode = "\x1b[?9001$p"
+)
+
+// Deprecated: use [SetWin32InputMode], [ResetWin32InputMode], and
+// [RequestWin32InputMode] instead.
+const (
+	EnableWin32Input  = "\x1b[?9001h"
+	DisableWin32Input = "\x1b[?9001l"
+	RequestWin32Input = "\x1b[?9001$p"
+)
diff --git a/vendor/github.com/charmbracelet/x/ansi/modes.go b/vendor/github.com/charmbracelet/x/ansi/modes.go
new file mode 100644
index 0000000..1bec5bc
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/modes.go
@@ -0,0 +1,71 @@
+package ansi
+
+// Modes represents the terminal modes that can be set or reset. By default,
+// all modes are [ModeNotRecognized].
+type Modes map[Mode]ModeSetting
+
+// NewModes creates a new Modes map. By default, all modes are
+// [ModeNotRecognized].
+func NewModes() Modes {
+	return make(Modes)
+}
+
+// Get returns the setting of a terminal mode. If the mode is not set, it
+// returns [ModeNotRecognized].
+func (m Modes) Get(mode Mode) ModeSetting {
+	return m[mode]
+}
+
+// Delete deletes a terminal mode. This has the same effect as setting the mode
+// to [ModeNotRecognized].
+func (m Modes) Delete(mode Mode) {
+	delete(m, mode)
+}
+
+// Set sets a terminal mode to [ModeSet].
+func (m Modes) Set(modes ...Mode) {
+	for _, mode := range modes {
+		m[mode] = ModeSet
+	}
+}
+
+// PermanentlySet sets a terminal mode to [ModePermanentlySet].
+func (m Modes) PermanentlySet(modes ...Mode) {
+	for _, mode := range modes {
+		m[mode] = ModePermanentlySet
+	}
+}
+
+// Reset sets a terminal mode to [ModeReset].
+func (m Modes) Reset(modes ...Mode) {
+	for _, mode := range modes {
+		m[mode] = ModeReset
+	}
+}
+
+// PermanentlyReset sets a terminal mode to [ModePermanentlyReset].
+func (m Modes) PermanentlyReset(modes ...Mode) {
+	for _, mode := range modes {
+		m[mode] = ModePermanentlyReset
+	}
+}
+
+// IsSet returns true if the mode is set to [ModeSet] or [ModePermanentlySet].
+func (m Modes) IsSet(mode Mode) bool {
+	return m[mode].IsSet()
+}
+
+// IsPermanentlySet returns true if the mode is set to [ModePermanentlySet].
+func (m Modes) IsPermanentlySet(mode Mode) bool {
+	return m[mode].IsPermanentlySet()
+}
+
+// IsReset returns true if the mode is set to [ModeReset] or [ModePermanentlyReset].
+func (m Modes) IsReset(mode Mode) bool {
+	return m[mode].IsReset()
+}
+
+// IsPermanentlyReset returns true if the mode is set to [ModePermanentlyReset].
+func (m Modes) IsPermanentlyReset(mode Mode) bool {
+	return m[mode].IsPermanentlyReset()
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/mouse.go b/vendor/github.com/charmbracelet/x/ansi/mouse.go
new file mode 100644
index 0000000..95b0127
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/mouse.go
@@ -0,0 +1,172 @@
+package ansi
+
+import (
+	"fmt"
+)
+
+// MouseButton represents the button that was pressed during a mouse message.
+type MouseButton byte
+
+// Mouse event buttons
+//
+// This is based on X11 mouse button codes.
+//
+//	1 = left button
+//	2 = middle button (pressing the scroll wheel)
+//	3 = right button
+//	4 = turn scroll wheel up
+//	5 = turn scroll wheel down
+//	6 = push scroll wheel left
+//	7 = push scroll wheel right
+//	8 = 4th button (aka browser backward button)
+//	9 = 5th button (aka browser forward button)
+//	10
+//	11
+//
+// Other buttons are not supported.
+const (
+	MouseNone MouseButton = iota
+	MouseButton1
+	MouseButton2
+	MouseButton3
+	MouseButton4
+	MouseButton5
+	MouseButton6
+	MouseButton7
+	MouseButton8
+	MouseButton9
+	MouseButton10
+	MouseButton11
+
+	MouseLeft       = MouseButton1
+	MouseMiddle     = MouseButton2
+	MouseRight      = MouseButton3
+	MouseWheelUp    = MouseButton4
+	MouseWheelDown  = MouseButton5
+	MouseWheelLeft  = MouseButton6
+	MouseWheelRight = MouseButton7
+	MouseBackward   = MouseButton8
+	MouseForward    = MouseButton9
+	MouseRelease    = MouseNone
+)
+
+var mouseButtons = map[MouseButton]string{
+	MouseNone:       "none",
+	MouseLeft:       "left",
+	MouseMiddle:     "middle",
+	MouseRight:      "right",
+	MouseWheelUp:    "wheelup",
+	MouseWheelDown:  "wheeldown",
+	MouseWheelLeft:  "wheelleft",
+	MouseWheelRight: "wheelright",
+	MouseBackward:   "backward",
+	MouseForward:    "forward",
+	MouseButton10:   "button10",
+	MouseButton11:   "button11",
+}
+
+// String returns a string representation of the mouse button.
+func (b MouseButton) String() string {
+	return mouseButtons[b]
+}
+
+// EncodeMouseButton returns a byte representing a mouse button.
+// The button is a bitmask of the following leftmost values:
+//
+//   - The first two bits are the button number:
+//     0 = left button, wheel up, or button no. 8 aka (backwards)
+//     1 = middle button, wheel down, or button no. 9 aka (forwards)
+//     2 = right button, wheel left, or button no. 10
+//     3 = release event, wheel right, or button no. 11
+//
+//   - The third bit indicates whether the shift key was pressed.
+//
+//   - The fourth bit indicates the alt key was pressed.
+//
+//   - The fifth bit indicates the control key was pressed.
+//
+//   - The sixth bit indicates motion events. Combined with button number 3, i.e.
+//     release event, it represents a drag event.
+//
+//   - The seventh bit indicates a wheel event.
+//
+//   - The eighth bit indicates additional buttons.
+//
+// If button is [MouseNone], and motion is false, this returns a release event.
+// If button is undefined, this function returns 0xff.
+func EncodeMouseButton(b MouseButton, motion, shift, alt, ctrl bool) (m byte) {
+	// mouse bit shifts
+	const (
+		bitShift  = 0b0000_0100
+		bitAlt    = 0b0000_1000
+		bitCtrl   = 0b0001_0000
+		bitMotion = 0b0010_0000
+		bitWheel  = 0b0100_0000
+		bitAdd    = 0b1000_0000 // additional buttons 8-11
+
+		bitsMask = 0b0000_0011
+	)
+
+	if b == MouseNone {
+		m = bitsMask
+	} else if b >= MouseLeft && b <= MouseRight {
+		m = byte(b - MouseLeft)
+	} else if b >= MouseWheelUp && b <= MouseWheelRight {
+		m = byte(b - MouseWheelUp)
+		m |= bitWheel
+	} else if b >= MouseBackward && b <= MouseButton11 {
+		m = byte(b - MouseBackward)
+		m |= bitAdd
+	} else {
+		m = 0xff // invalid button
+	}
+
+	if shift {
+		m |= bitShift
+	}
+	if alt {
+		m |= bitAlt
+	}
+	if ctrl {
+		m |= bitCtrl
+	}
+	if motion {
+		m |= bitMotion
+	}
+
+	return
+}
+
+// x10Offset is the offset for X10 mouse events.
+// See https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#Mouse%20Tracking
+const x10Offset = 32
+
+// MouseX10 returns an escape sequence representing a mouse event in X10 mode.
+// Note that this requires the terminal support X10 mouse modes.
+//
+//	CSI M Cb Cx Cy
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#Mouse%20Tracking
+func MouseX10(b byte, x, y int) string {
+	return "\x1b[M" + string(b+x10Offset) + string(byte(x)+x10Offset+1) + string(byte(y)+x10Offset+1)
+}
+
+// MouseSgr returns an escape sequence representing a mouse event in SGR mode.
+//
+//	CSI < Cb ; Cx ; Cy M
+//	CSI < Cb ; Cx ; Cy m (release)
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#Mouse%20Tracking
+func MouseSgr(b byte, x, y int, release bool) string {
+	s := 'M'
+	if release {
+		s = 'm'
+	}
+	if x < 0 {
+		x = -x
+	}
+	if y < 0 {
+		y = -y
+	}
+	return fmt.Sprintf("\x1b[<%d;%d;%d%c", b, x+1, y+1, s)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/notification.go b/vendor/github.com/charmbracelet/x/ansi/notification.go
new file mode 100644
index 0000000..c712f34
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/notification.go
@@ -0,0 +1,13 @@
+package ansi
+
+// Notify sends a desktop notification using iTerm's OSC 9.
+//
+//	OSC 9 ; Mc ST
+//	OSC 9 ; Mc BEL
+//
+// Where Mc is the notification body.
+//
+// See: https://iterm2.com/documentation-escape-codes.html
+func Notify(s string) string {
+	return "\x1b]9;" + s + "\x07"
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser.go b/vendor/github.com/charmbracelet/x/ansi/parser.go
new file mode 100644
index 0000000..882e1ed
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser.go
@@ -0,0 +1,417 @@
+package ansi
+
+import (
+	"unicode/utf8"
+	"unsafe"
+
+	"github.com/charmbracelet/x/ansi/parser"
+)
+
+// Parser represents a DEC ANSI compatible sequence parser.
+//
+// It uses a state machine to parse ANSI escape sequences and control
+// characters. The parser is designed to be used with a terminal emulator or
+// similar application that needs to parse ANSI escape sequences and control
+// characters.
+// See package [parser] for more information.
+//
+//go:generate go run ./gen.go
+type Parser struct {
+	handler Handler
+
+	// params contains the raw parameters of the sequence.
+	// These parameters used when constructing CSI and DCS sequences.
+	params []int
+
+	// data contains the raw data of the sequence.
+	// These data used when constructing OSC, DCS, SOS, PM, and APC sequences.
+	data []byte
+
+	// dataLen keeps track of the length of the data buffer.
+	// If dataLen is -1, the data buffer is unlimited and will grow as needed.
+	// Otherwise, dataLen is limited by the size of the data buffer.
+	dataLen int
+
+	// paramsLen keeps track of the number of parameters.
+	// This is limited by the size of the params buffer.
+	//
+	// This is also used when collecting UTF-8 runes to keep track of the
+	// number of rune bytes collected.
+	paramsLen int
+
+	// cmd contains the raw command along with the private prefix and
+	// intermediate bytes of the sequence.
+	// The first lower byte contains the command byte, the next byte contains
+	// the private prefix, and the next byte contains the intermediate byte.
+	//
+	// This is also used when collecting UTF-8 runes treating it as a slice of
+	// 4 bytes.
+	cmd int
+
+	// state is the current state of the parser.
+	state byte
+}
+
+// NewParser returns a new parser with the default settings.
+// The [Parser] uses a default size of 32 for the parameters and 64KB for the
+// data buffer. Use [Parser.SetParamsSize] and [Parser.SetDataSize] to set the
+// size of the parameters and data buffer respectively.
+func NewParser() *Parser {
+	p := new(Parser)
+	p.SetParamsSize(parser.MaxParamsSize)
+	p.SetDataSize(1024 * 64) // 64KB data buffer
+	return p
+}
+
+// SetParamsSize sets the size of the parameters buffer.
+// This is used when constructing CSI and DCS sequences.
+func (p *Parser) SetParamsSize(size int) {
+	p.params = make([]int, size)
+}
+
+// SetDataSize sets the size of the data buffer.
+// This is used when constructing OSC, DCS, SOS, PM, and APC sequences.
+// If size is less than or equal to 0, the data buffer is unlimited and will
+// grow as needed.
+func (p *Parser) SetDataSize(size int) {
+	if size <= 0 {
+		size = 0
+		p.dataLen = -1
+	}
+	p.data = make([]byte, size)
+}
+
+// Params returns the list of parsed packed parameters.
+func (p *Parser) Params() Params {
+	return unsafe.Slice((*Param)(unsafe.Pointer(&p.params[0])), p.paramsLen)
+}
+
+// Param returns the parameter at the given index and falls back to the default
+// value if the parameter is missing. If the index is out of bounds, it returns
+// the default value and false.
+func (p *Parser) Param(i, def int) (int, bool) {
+	if i < 0 || i >= p.paramsLen {
+		return def, false
+	}
+	return Param(p.params[i]).Param(def), true
+}
+
+// Command returns the packed command of the last dispatched sequence. Use
+// [Cmd] to unpack the command.
+func (p *Parser) Command() int {
+	return p.cmd
+}
+
+// Rune returns the last dispatched sequence as a rune.
+func (p *Parser) Rune() rune {
+	rw := utf8ByteLen(byte(p.cmd & 0xff))
+	if rw == -1 {
+		return utf8.RuneError
+	}
+	r, _ := utf8.DecodeRune((*[utf8.UTFMax]byte)(unsafe.Pointer(&p.cmd))[:rw])
+	return r
+}
+
+// Control returns the last dispatched sequence as a control code.
+func (p *Parser) Control() byte {
+	return byte(p.cmd & 0xff)
+}
+
+// Data returns the raw data of the last dispatched sequence.
+func (p *Parser) Data() []byte {
+	return p.data[:p.dataLen]
+}
+
+// Reset resets the parser to its initial state.
+func (p *Parser) Reset() {
+	p.clear()
+	p.state = parser.GroundState
+}
+
+// clear clears the parser parameters and command.
+func (p *Parser) clear() {
+	if len(p.params) > 0 {
+		p.params[0] = parser.MissingParam
+	}
+	p.paramsLen = 0
+	p.cmd = 0
+}
+
+// State returns the current state of the parser.
+func (p *Parser) State() parser.State {
+	return p.state
+}
+
+// StateName returns the name of the current state.
+func (p *Parser) StateName() string {
+	return parser.StateNames[p.state]
+}
+
+// Parse parses the given dispatcher and byte buffer.
+// Deprecated: Loop over the buffer and call [Parser.Advance] instead.
+func (p *Parser) Parse(b []byte) {
+	for i := 0; i < len(b); i++ {
+		p.Advance(b[i])
+	}
+}
+
+// Advance advances the parser using the given byte. It	returns the action
+// performed by the parser.
+func (p *Parser) Advance(b byte) parser.Action {
+	switch p.state {
+	case parser.Utf8State:
+		// We handle UTF-8 here.
+		return p.advanceUtf8(b)
+	default:
+		return p.advance(b)
+	}
+}
+
+func (p *Parser) collectRune(b byte) {
+	if p.paramsLen >= utf8.UTFMax {
+		return
+	}
+
+	shift := p.paramsLen * 8
+	p.cmd &^= 0xff << shift
+	p.cmd |= int(b) << shift
+	p.paramsLen++
+}
+
+func (p *Parser) advanceUtf8(b byte) parser.Action {
+	// Collect UTF-8 rune bytes.
+	p.collectRune(b)
+	rw := utf8ByteLen(byte(p.cmd & 0xff))
+	if rw == -1 {
+		// We panic here because the first byte comes from the state machine,
+		// if this panics, it means there is a bug in the state machine!
+		panic("invalid rune") // unreachable
+	}
+
+	if p.paramsLen < rw {
+		return parser.CollectAction
+	}
+
+	// We have enough bytes to decode the rune using unsafe
+	if p.handler.Print != nil {
+		p.handler.Print(p.Rune())
+	}
+
+	p.state = parser.GroundState
+	p.paramsLen = 0
+
+	return parser.PrintAction
+}
+
+func (p *Parser) advance(b byte) parser.Action {
+	state, action := parser.Table.Transition(p.state, b)
+
+	// We need to clear the parser state if the state changes from EscapeState.
+	// This is because when we enter the EscapeState, we don't get a chance to
+	// clear the parser state. For example, when a sequence terminates with a
+	// ST (\x1b\\ or \x9c), we dispatch the current sequence and transition to
+	// EscapeState. However, the parser state is not cleared in this case and
+	// we need to clear it here before dispatching the esc sequence.
+	if p.state != state {
+		if p.state == parser.EscapeState {
+			p.performAction(parser.ClearAction, state, b)
+		}
+		if action == parser.PutAction &&
+			p.state == parser.DcsEntryState && state == parser.DcsStringState {
+			// XXX: This is a special case where we need to start collecting
+			// non-string parameterized data i.e. doesn't follow the ECMA-48 §
+			// 5.4.1 string parameters format.
+			p.performAction(parser.StartAction, state, 0)
+		}
+	}
+
+	// Handle special cases
+	switch {
+	case b == ESC && p.state == parser.EscapeState:
+		// Two ESCs in a row
+		p.performAction(parser.ExecuteAction, state, b)
+	default:
+		p.performAction(action, state, b)
+	}
+
+	p.state = state
+
+	return action
+}
+
+func (p *Parser) parseStringCmd() {
+	// Try to parse the command
+	datalen := len(p.data)
+	if p.dataLen >= 0 {
+		datalen = p.dataLen
+	}
+	for i := 0; i < datalen; i++ {
+		d := p.data[i]
+		if d < '0' || d > '9' {
+			break
+		}
+		if p.cmd == parser.MissingCommand {
+			p.cmd = 0
+		}
+		p.cmd *= 10
+		p.cmd += int(d - '0')
+	}
+}
+
+func (p *Parser) performAction(action parser.Action, state parser.State, b byte) {
+	switch action {
+	case parser.IgnoreAction:
+		break
+
+	case parser.ClearAction:
+		p.clear()
+
+	case parser.PrintAction:
+		p.cmd = int(b)
+		if p.handler.Print != nil {
+			p.handler.Print(rune(b))
+		}
+
+	case parser.ExecuteAction:
+		p.cmd = int(b)
+		if p.handler.Execute != nil {
+			p.handler.Execute(b)
+		}
+
+	case parser.PrefixAction:
+		// Collect private prefix
+		// we only store the last prefix
+		p.cmd &^= 0xff << parser.PrefixShift
+		p.cmd |= int(b) << parser.PrefixShift
+
+	case parser.CollectAction:
+		if state == parser.Utf8State {
+			// Reset the UTF-8 counter
+			p.paramsLen = 0
+			p.collectRune(b)
+		} else {
+			// Collect intermediate bytes
+			// we only store the last intermediate byte
+			p.cmd &^= 0xff << parser.IntermedShift
+			p.cmd |= int(b) << parser.IntermedShift
+		}
+
+	case parser.ParamAction:
+		// Collect parameters
+		if p.paramsLen >= len(p.params) {
+			break
+		}
+
+		if b >= '0' && b <= '9' {
+			if p.params[p.paramsLen] == parser.MissingParam {
+				p.params[p.paramsLen] = 0
+			}
+
+			p.params[p.paramsLen] *= 10
+			p.params[p.paramsLen] += int(b - '0')
+		}
+
+		if b == ':' {
+			p.params[p.paramsLen] |= parser.HasMoreFlag
+		}
+
+		if b == ';' || b == ':' {
+			p.paramsLen++
+			if p.paramsLen < len(p.params) {
+				p.params[p.paramsLen] = parser.MissingParam
+			}
+		}
+
+	case parser.StartAction:
+		if p.dataLen < 0 && p.data != nil {
+			p.data = p.data[:0]
+		} else {
+			p.dataLen = 0
+		}
+		if p.state >= parser.DcsEntryState && p.state <= parser.DcsStringState {
+			// Collect the command byte for DCS
+			p.cmd |= int(b)
+		} else {
+			p.cmd = parser.MissingCommand
+		}
+
+	case parser.PutAction:
+		switch p.state {
+		case parser.OscStringState:
+			if b == ';' && p.cmd == parser.MissingCommand {
+				p.parseStringCmd()
+			}
+		}
+
+		if p.dataLen < 0 {
+			p.data = append(p.data, b)
+		} else {
+			if p.dataLen < len(p.data) {
+				p.data[p.dataLen] = b
+				p.dataLen++
+			}
+		}
+
+	case parser.DispatchAction:
+		// Increment the last parameter
+		if p.paramsLen > 0 && p.paramsLen < len(p.params)-1 ||
+			p.paramsLen == 0 && len(p.params) > 0 && p.params[0] != parser.MissingParam {
+			p.paramsLen++
+		}
+
+		if p.state == parser.OscStringState && p.cmd == parser.MissingCommand {
+			// Ensure we have a command for OSC
+			p.parseStringCmd()
+		}
+
+		data := p.data
+		if p.dataLen >= 0 {
+			data = data[:p.dataLen]
+		}
+		switch p.state {
+		case parser.CsiEntryState, parser.CsiParamState, parser.CsiIntermediateState:
+			p.cmd |= int(b)
+			if p.handler.HandleCsi != nil {
+				p.handler.HandleCsi(Cmd(p.cmd), p.Params())
+			}
+		case parser.EscapeState, parser.EscapeIntermediateState:
+			p.cmd |= int(b)
+			if p.handler.HandleEsc != nil {
+				p.handler.HandleEsc(Cmd(p.cmd))
+			}
+		case parser.DcsEntryState, parser.DcsParamState, parser.DcsIntermediateState, parser.DcsStringState:
+			if p.handler.HandleDcs != nil {
+				p.handler.HandleDcs(Cmd(p.cmd), p.Params(), data)
+			}
+		case parser.OscStringState:
+			if p.handler.HandleOsc != nil {
+				p.handler.HandleOsc(p.cmd, data)
+			}
+		case parser.SosStringState:
+			if p.handler.HandleSos != nil {
+				p.handler.HandleSos(data)
+			}
+		case parser.PmStringState:
+			if p.handler.HandlePm != nil {
+				p.handler.HandlePm(data)
+			}
+		case parser.ApcStringState:
+			if p.handler.HandleApc != nil {
+				p.handler.HandleApc(data)
+			}
+		}
+	}
+}
+
+func utf8ByteLen(b byte) int {
+	if b <= 0b0111_1111 { // 0x00-0x7F
+		return 1
+	} else if b >= 0b1100_0000 && b <= 0b1101_1111 { // 0xC0-0xDF
+		return 2
+	} else if b >= 0b1110_0000 && b <= 0b1110_1111 { // 0xE0-0xEF
+		return 3
+	} else if b >= 0b1111_0000 && b <= 0b1111_0111 { // 0xF0-0xF7
+		return 4
+	}
+	return -1
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser/const.go b/vendor/github.com/charmbracelet/x/ansi/parser/const.go
new file mode 100644
index 0000000..d62dbf3
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser/const.go
@@ -0,0 +1,78 @@
+package parser
+
+// Action is a DEC ANSI parser action.
+type Action = byte
+
+// These are the actions that the parser can take.
+const (
+	NoneAction Action = iota
+	ClearAction
+	CollectAction
+	PrefixAction
+	DispatchAction
+	ExecuteAction
+	StartAction // Start of a data string
+	PutAction   // Put into the data string
+	ParamAction
+	PrintAction
+
+	IgnoreAction = NoneAction
+)
+
+// nolint: unused
+var ActionNames = []string{
+	"NoneAction",
+	"ClearAction",
+	"CollectAction",
+	"PrefixAction",
+	"DispatchAction",
+	"ExecuteAction",
+	"StartAction",
+	"PutAction",
+	"ParamAction",
+	"PrintAction",
+}
+
+// State is a DEC ANSI parser state.
+type State = byte
+
+// These are the states that the parser can be in.
+const (
+	GroundState State = iota
+	CsiEntryState
+	CsiIntermediateState
+	CsiParamState
+	DcsEntryState
+	DcsIntermediateState
+	DcsParamState
+	DcsStringState
+	EscapeState
+	EscapeIntermediateState
+	OscStringState
+	SosStringState
+	PmStringState
+	ApcStringState
+
+	// Utf8State is not part of the DEC ANSI standard. It is used to handle
+	// UTF-8 sequences.
+	Utf8State
+)
+
+// nolint: unused
+var StateNames = []string{
+	"GroundState",
+	"CsiEntryState",
+	"CsiIntermediateState",
+	"CsiParamState",
+	"DcsEntryState",
+	"DcsIntermediateState",
+	"DcsParamState",
+	"DcsStringState",
+	"EscapeState",
+	"EscapeIntermediateState",
+	"OscStringState",
+	"SosStringState",
+	"PmStringState",
+	"ApcStringState",
+	"Utf8State",
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser/seq.go b/vendor/github.com/charmbracelet/x/ansi/parser/seq.go
new file mode 100644
index 0000000..29f491d
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser/seq.go
@@ -0,0 +1,136 @@
+package parser
+
+import "math"
+
+// Shift and masks for sequence parameters and intermediates.
+const (
+	PrefixShift    = 8
+	IntermedShift  = 16
+	FinalMask      = 0xff
+	HasMoreFlag    = math.MinInt32
+	ParamMask      = ^HasMoreFlag
+	MissingParam   = ParamMask
+	MissingCommand = MissingParam
+	MaxParam       = math.MaxUint16 // the maximum value a parameter can have
+)
+
+const (
+	// MaxParamsSize is the maximum number of parameters a sequence can have.
+	MaxParamsSize = 32
+
+	// DefaultParamValue is the default value used for missing parameters.
+	DefaultParamValue = 0
+)
+
+// Prefix returns the prefix byte of the sequence.
+// This is always gonna be one of the following '<' '=' '>' '?' and in the
+// range of 0x3C-0x3F.
+// Zero is returned if the sequence does not have a prefix.
+func Prefix(cmd int) int {
+	return (cmd >> PrefixShift) & FinalMask
+}
+
+// Intermediate returns the intermediate byte of the sequence.
+// An intermediate byte is in the range of 0x20-0x2F. This includes these
+// characters from ' ', '!', '"', '#', '$', '%', '&', ”', '(', ')', '*', '+',
+// ',', '-', '.', '/'.
+// Zero is returned if the sequence does not have an intermediate byte.
+func Intermediate(cmd int) int {
+	return (cmd >> IntermedShift) & FinalMask
+}
+
+// Command returns the command byte of the CSI sequence.
+func Command(cmd int) int {
+	return cmd & FinalMask
+}
+
+// Param returns the parameter at the given index.
+// It returns -1 if the parameter does not exist.
+func Param(params []int, i int) int {
+	if len(params) == 0 || i < 0 || i >= len(params) {
+		return -1
+	}
+
+	p := params[i] & ParamMask
+	if p == MissingParam {
+		return -1
+	}
+
+	return p
+}
+
+// HasMore returns true if the parameter has more sub-parameters.
+func HasMore(params []int, i int) bool {
+	if len(params) == 0 || i >= len(params) {
+		return false
+	}
+
+	return params[i]&HasMoreFlag != 0
+}
+
+// Subparams returns the sub-parameters of the given parameter.
+// It returns nil if the parameter does not exist.
+func Subparams(params []int, i int) []int {
+	if len(params) == 0 || i < 0 || i >= len(params) {
+		return nil
+	}
+
+	// Count the number of parameters before the given parameter index.
+	var count int
+	var j int
+	for j = 0; j < len(params); j++ {
+		if count == i {
+			break
+		}
+		if !HasMore(params, j) {
+			count++
+		}
+	}
+
+	if count > i || j >= len(params) {
+		return nil
+	}
+
+	var subs []int
+	for ; j < len(params); j++ {
+		if !HasMore(params, j) {
+			break
+		}
+		p := Param(params, j)
+		if p == -1 {
+			p = DefaultParamValue
+		}
+		subs = append(subs, p)
+	}
+
+	p := Param(params, j)
+	if p == -1 {
+		p = DefaultParamValue
+	}
+
+	return append(subs, p)
+}
+
+// Len returns the number of parameters in the sequence.
+// This will return the number of parameters in the sequence, excluding any
+// sub-parameters.
+func Len(params []int) int {
+	var n int
+	for i := 0; i < len(params); i++ {
+		if !HasMore(params, i) {
+			n++
+		}
+	}
+	return n
+}
+
+// Range iterates over the parameters of the sequence and calls the given
+// function for each parameter.
+// The function should return false to stop the iteration.
+func Range(params []int, fn func(i int, param int, hasMore bool) bool) {
+	for i := 0; i < len(params); i++ {
+		if !fn(i, Param(params, i), HasMore(params, i)) {
+			break
+		}
+	}
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser/transition_table.go b/vendor/github.com/charmbracelet/x/ansi/parser/transition_table.go
new file mode 100644
index 0000000..558a5ea
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser/transition_table.go
@@ -0,0 +1,273 @@
+package parser
+
+// Table values are generated like this:
+//
+//	index:  currentState << IndexStateShift | charCode
+//	value:  action << TransitionActionShift | nextState
+const (
+	TransitionActionShift = 4
+	TransitionStateMask   = 15
+	IndexStateShift       = 8
+
+	// DefaultTableSize is the default size of the transition table.
+	DefaultTableSize = 4096
+)
+
+// Table is a DEC ANSI transition table.
+var Table = GenerateTransitionTable()
+
+// TransitionTable is a DEC ANSI transition table.
+// https://vt100.net/emu/dec_ansi_parser
+type TransitionTable []byte
+
+// NewTransitionTable returns a new DEC ANSI transition table.
+func NewTransitionTable(size int) TransitionTable {
+	if size <= 0 {
+		size = DefaultTableSize
+	}
+	return TransitionTable(make([]byte, size))
+}
+
+// SetDefault sets default transition.
+func (t TransitionTable) SetDefault(action Action, state State) {
+	for i := 0; i < len(t); i++ {
+		t[i] = action<<TransitionActionShift | state
+	}
+}
+
+// AddOne adds a transition.
+func (t TransitionTable) AddOne(code byte, state State, action Action, next State) {
+	idx := int(state)<<IndexStateShift | int(code)
+	value := action<<TransitionActionShift | next
+	t[idx] = value
+}
+
+// AddMany adds many transitions.
+func (t TransitionTable) AddMany(codes []byte, state State, action Action, next State) {
+	for _, code := range codes {
+		t.AddOne(code, state, action, next)
+	}
+}
+
+// AddRange adds a range of transitions.
+func (t TransitionTable) AddRange(start, end byte, state State, action Action, next State) {
+	for i := int(start); i <= int(end); i++ {
+		t.AddOne(byte(i), state, action, next)
+	}
+}
+
+// Transition returns the next state and action for the given state and byte.
+func (t TransitionTable) Transition(state State, code byte) (State, Action) {
+	index := int(state)<<IndexStateShift | int(code)
+	value := t[index]
+	return value & TransitionStateMask, value >> TransitionActionShift
+}
+
+// byte range macro
+func r(start, end byte) []byte {
+	var a []byte
+	for i := int(start); i <= int(end); i++ {
+		a = append(a, byte(i))
+	}
+	return a
+}
+
+// GenerateTransitionTable generates a DEC ANSI transition table compatible
+// with the VT500-series of terminals. This implementation includes a few
+// modifications that include:
+//   - A new Utf8State is introduced to handle UTF8 sequences.
+//   - Osc and Dcs data accept UTF8 sequences by extending the printable range
+//     to 0xFF and 0xFE respectively.
+//   - We don't ignore 0x3A (':') when building Csi and Dcs parameters and
+//     instead use it to denote sub-parameters.
+//   - Support dispatching SosPmApc sequences.
+//   - The DEL (0x7F) character is executed in the Ground state.
+//   - The DEL (0x7F) character is collected in the DcsPassthrough string state.
+//   - The ST C1 control character (0x9C) is executed and not ignored.
+func GenerateTransitionTable() TransitionTable {
+	table := NewTransitionTable(DefaultTableSize)
+	table.SetDefault(NoneAction, GroundState)
+
+	// Anywhere
+	for _, state := range r(GroundState, Utf8State) {
+		// Anywhere -> Ground
+		table.AddMany([]byte{0x18, 0x1a, 0x99, 0x9a}, state, ExecuteAction, GroundState)
+		table.AddRange(0x80, 0x8F, state, ExecuteAction, GroundState)
+		table.AddRange(0x90, 0x97, state, ExecuteAction, GroundState)
+		table.AddOne(0x9C, state, ExecuteAction, GroundState)
+		// Anywhere -> Escape
+		table.AddOne(0x1B, state, ClearAction, EscapeState)
+		// Anywhere -> SosStringState
+		table.AddOne(0x98, state, StartAction, SosStringState)
+		// Anywhere -> PmStringState
+		table.AddOne(0x9E, state, StartAction, PmStringState)
+		// Anywhere -> ApcStringState
+		table.AddOne(0x9F, state, StartAction, ApcStringState)
+		// Anywhere -> CsiEntry
+		table.AddOne(0x9B, state, ClearAction, CsiEntryState)
+		// Anywhere -> DcsEntry
+		table.AddOne(0x90, state, ClearAction, DcsEntryState)
+		// Anywhere -> OscString
+		table.AddOne(0x9D, state, StartAction, OscStringState)
+		// Anywhere -> Utf8
+		table.AddRange(0xC2, 0xDF, state, CollectAction, Utf8State) // UTF8 2 byte sequence
+		table.AddRange(0xE0, 0xEF, state, CollectAction, Utf8State) // UTF8 3 byte sequence
+		table.AddRange(0xF0, 0xF4, state, CollectAction, Utf8State) // UTF8 4 byte sequence
+	}
+
+	// Ground
+	table.AddRange(0x00, 0x17, GroundState, ExecuteAction, GroundState)
+	table.AddOne(0x19, GroundState, ExecuteAction, GroundState)
+	table.AddRange(0x1C, 0x1F, GroundState, ExecuteAction, GroundState)
+	table.AddRange(0x20, 0x7E, GroundState, PrintAction, GroundState)
+	table.AddOne(0x7F, GroundState, ExecuteAction, GroundState)
+
+	// EscapeIntermediate
+	table.AddRange(0x00, 0x17, EscapeIntermediateState, ExecuteAction, EscapeIntermediateState)
+	table.AddOne(0x19, EscapeIntermediateState, ExecuteAction, EscapeIntermediateState)
+	table.AddRange(0x1C, 0x1F, EscapeIntermediateState, ExecuteAction, EscapeIntermediateState)
+	table.AddRange(0x20, 0x2F, EscapeIntermediateState, CollectAction, EscapeIntermediateState)
+	table.AddOne(0x7F, EscapeIntermediateState, IgnoreAction, EscapeIntermediateState)
+	// EscapeIntermediate -> Ground
+	table.AddRange(0x30, 0x7E, EscapeIntermediateState, DispatchAction, GroundState)
+
+	// Escape
+	table.AddRange(0x00, 0x17, EscapeState, ExecuteAction, EscapeState)
+	table.AddOne(0x19, EscapeState, ExecuteAction, EscapeState)
+	table.AddRange(0x1C, 0x1F, EscapeState, ExecuteAction, EscapeState)
+	table.AddOne(0x7F, EscapeState, IgnoreAction, EscapeState)
+	// Escape -> Ground
+	table.AddRange(0x30, 0x4F, EscapeState, DispatchAction, GroundState)
+	table.AddRange(0x51, 0x57, EscapeState, DispatchAction, GroundState)
+	table.AddOne(0x59, EscapeState, DispatchAction, GroundState)
+	table.AddOne(0x5A, EscapeState, DispatchAction, GroundState)
+	table.AddOne(0x5C, EscapeState, DispatchAction, GroundState)
+	table.AddRange(0x60, 0x7E, EscapeState, DispatchAction, GroundState)
+	// Escape -> Escape_intermediate
+	table.AddRange(0x20, 0x2F, EscapeState, CollectAction, EscapeIntermediateState)
+	// Escape -> Sos_pm_apc_string
+	table.AddOne('X', EscapeState, StartAction, SosStringState) // SOS
+	table.AddOne('^', EscapeState, StartAction, PmStringState)  // PM
+	table.AddOne('_', EscapeState, StartAction, ApcStringState) // APC
+	// Escape -> Dcs_entry
+	table.AddOne('P', EscapeState, ClearAction, DcsEntryState)
+	// Escape -> Csi_entry
+	table.AddOne('[', EscapeState, ClearAction, CsiEntryState)
+	// Escape -> Osc_string
+	table.AddOne(']', EscapeState, StartAction, OscStringState)
+
+	// Sos_pm_apc_string
+	for _, state := range r(SosStringState, ApcStringState) {
+		table.AddRange(0x00, 0x17, state, PutAction, state)
+		table.AddOne(0x19, state, PutAction, state)
+		table.AddRange(0x1C, 0x1F, state, PutAction, state)
+		table.AddRange(0x20, 0x7F, state, PutAction, state)
+		// ESC, ST, CAN, and SUB terminate the sequence
+		table.AddOne(0x1B, state, DispatchAction, EscapeState)
+		table.AddOne(0x9C, state, DispatchAction, GroundState)
+		table.AddMany([]byte{0x18, 0x1A}, state, IgnoreAction, GroundState)
+	}
+
+	// Dcs_entry
+	table.AddRange(0x00, 0x07, DcsEntryState, IgnoreAction, DcsEntryState)
+	table.AddRange(0x0E, 0x17, DcsEntryState, IgnoreAction, DcsEntryState)
+	table.AddOne(0x19, DcsEntryState, IgnoreAction, DcsEntryState)
+	table.AddRange(0x1C, 0x1F, DcsEntryState, IgnoreAction, DcsEntryState)
+	table.AddOne(0x7F, DcsEntryState, IgnoreAction, DcsEntryState)
+	// Dcs_entry -> Dcs_intermediate
+	table.AddRange(0x20, 0x2F, DcsEntryState, CollectAction, DcsIntermediateState)
+	// Dcs_entry -> Dcs_param
+	table.AddRange(0x30, 0x3B, DcsEntryState, ParamAction, DcsParamState)
+	table.AddRange(0x3C, 0x3F, DcsEntryState, PrefixAction, DcsParamState)
+	// Dcs_entry -> Dcs_passthrough
+	table.AddRange(0x08, 0x0D, DcsEntryState, PutAction, DcsStringState) // Follows ECMA-48 § 8.3.27
+	// XXX: allows passing ESC (not a ECMA-48 standard) this to allow for
+	// passthrough of ANSI sequences like in Screen or Tmux passthrough mode.
+	table.AddOne(0x1B, DcsEntryState, PutAction, DcsStringState)
+	table.AddRange(0x40, 0x7E, DcsEntryState, StartAction, DcsStringState)
+
+	// Dcs_intermediate
+	table.AddRange(0x00, 0x17, DcsIntermediateState, IgnoreAction, DcsIntermediateState)
+	table.AddOne(0x19, DcsIntermediateState, IgnoreAction, DcsIntermediateState)
+	table.AddRange(0x1C, 0x1F, DcsIntermediateState, IgnoreAction, DcsIntermediateState)
+	table.AddRange(0x20, 0x2F, DcsIntermediateState, CollectAction, DcsIntermediateState)
+	table.AddOne(0x7F, DcsIntermediateState, IgnoreAction, DcsIntermediateState)
+	// Dcs_intermediate -> Dcs_passthrough
+	table.AddRange(0x30, 0x3F, DcsIntermediateState, StartAction, DcsStringState)
+	table.AddRange(0x40, 0x7E, DcsIntermediateState, StartAction, DcsStringState)
+
+	// Dcs_param
+	table.AddRange(0x00, 0x17, DcsParamState, IgnoreAction, DcsParamState)
+	table.AddOne(0x19, DcsParamState, IgnoreAction, DcsParamState)
+	table.AddRange(0x1C, 0x1F, DcsParamState, IgnoreAction, DcsParamState)
+	table.AddRange(0x30, 0x3B, DcsParamState, ParamAction, DcsParamState)
+	table.AddOne(0x7F, DcsParamState, IgnoreAction, DcsParamState)
+	table.AddRange(0x3C, 0x3F, DcsParamState, IgnoreAction, DcsParamState)
+	// Dcs_param -> Dcs_intermediate
+	table.AddRange(0x20, 0x2F, DcsParamState, CollectAction, DcsIntermediateState)
+	// Dcs_param -> Dcs_passthrough
+	table.AddRange(0x40, 0x7E, DcsParamState, StartAction, DcsStringState)
+
+	// Dcs_passthrough
+	table.AddRange(0x00, 0x17, DcsStringState, PutAction, DcsStringState)
+	table.AddOne(0x19, DcsStringState, PutAction, DcsStringState)
+	table.AddRange(0x1C, 0x1F, DcsStringState, PutAction, DcsStringState)
+	table.AddRange(0x20, 0x7E, DcsStringState, PutAction, DcsStringState)
+	table.AddOne(0x7F, DcsStringState, PutAction, DcsStringState)
+	table.AddRange(0x80, 0xFF, DcsStringState, PutAction, DcsStringState) // Allow Utf8 characters by extending the printable range from 0x7F to 0xFF
+	// ST, CAN, SUB, and ESC terminate the sequence
+	table.AddOne(0x1B, DcsStringState, DispatchAction, EscapeState)
+	table.AddOne(0x9C, DcsStringState, DispatchAction, GroundState)
+	table.AddMany([]byte{0x18, 0x1A}, DcsStringState, IgnoreAction, GroundState)
+
+	// Csi_param
+	table.AddRange(0x00, 0x17, CsiParamState, ExecuteAction, CsiParamState)
+	table.AddOne(0x19, CsiParamState, ExecuteAction, CsiParamState)
+	table.AddRange(0x1C, 0x1F, CsiParamState, ExecuteAction, CsiParamState)
+	table.AddRange(0x30, 0x3B, CsiParamState, ParamAction, CsiParamState)
+	table.AddOne(0x7F, CsiParamState, IgnoreAction, CsiParamState)
+	table.AddRange(0x3C, 0x3F, CsiParamState, IgnoreAction, CsiParamState)
+	// Csi_param -> Ground
+	table.AddRange(0x40, 0x7E, CsiParamState, DispatchAction, GroundState)
+	// Csi_param -> Csi_intermediate
+	table.AddRange(0x20, 0x2F, CsiParamState, CollectAction, CsiIntermediateState)
+
+	// Csi_intermediate
+	table.AddRange(0x00, 0x17, CsiIntermediateState, ExecuteAction, CsiIntermediateState)
+	table.AddOne(0x19, CsiIntermediateState, ExecuteAction, CsiIntermediateState)
+	table.AddRange(0x1C, 0x1F, CsiIntermediateState, ExecuteAction, CsiIntermediateState)
+	table.AddRange(0x20, 0x2F, CsiIntermediateState, CollectAction, CsiIntermediateState)
+	table.AddOne(0x7F, CsiIntermediateState, IgnoreAction, CsiIntermediateState)
+	// Csi_intermediate -> Ground
+	table.AddRange(0x40, 0x7E, CsiIntermediateState, DispatchAction, GroundState)
+	// Csi_intermediate -> Csi_ignore
+	table.AddRange(0x30, 0x3F, CsiIntermediateState, IgnoreAction, GroundState)
+
+	// Csi_entry
+	table.AddRange(0x00, 0x17, CsiEntryState, ExecuteAction, CsiEntryState)
+	table.AddOne(0x19, CsiEntryState, ExecuteAction, CsiEntryState)
+	table.AddRange(0x1C, 0x1F, CsiEntryState, ExecuteAction, CsiEntryState)
+	table.AddOne(0x7F, CsiEntryState, IgnoreAction, CsiEntryState)
+	// Csi_entry -> Ground
+	table.AddRange(0x40, 0x7E, CsiEntryState, DispatchAction, GroundState)
+	// Csi_entry -> Csi_intermediate
+	table.AddRange(0x20, 0x2F, CsiEntryState, CollectAction, CsiIntermediateState)
+	// Csi_entry -> Csi_param
+	table.AddRange(0x30, 0x3B, CsiEntryState, ParamAction, CsiParamState)
+	table.AddRange(0x3C, 0x3F, CsiEntryState, PrefixAction, CsiParamState)
+
+	// Osc_string
+	table.AddRange(0x00, 0x06, OscStringState, IgnoreAction, OscStringState)
+	table.AddRange(0x08, 0x17, OscStringState, IgnoreAction, OscStringState)
+	table.AddOne(0x19, OscStringState, IgnoreAction, OscStringState)
+	table.AddRange(0x1C, 0x1F, OscStringState, IgnoreAction, OscStringState)
+	table.AddRange(0x20, 0xFF, OscStringState, PutAction, OscStringState) // Allow Utf8 characters by extending the printable range from 0x7F to 0xFF
+
+	// ST, CAN, SUB, ESC, and BEL terminate the sequence
+	table.AddOne(0x1B, OscStringState, DispatchAction, EscapeState)
+	table.AddOne(0x07, OscStringState, DispatchAction, GroundState)
+	table.AddOne(0x9C, OscStringState, DispatchAction, GroundState)
+	table.AddMany([]byte{0x18, 0x1A}, OscStringState, IgnoreAction, GroundState)
+
+	return table
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser_decode.go b/vendor/github.com/charmbracelet/x/ansi/parser_decode.go
new file mode 100644
index 0000000..3e50473
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser_decode.go
@@ -0,0 +1,524 @@
+package ansi
+
+import (
+	"unicode/utf8"
+
+	"github.com/charmbracelet/x/ansi/parser"
+	"github.com/mattn/go-runewidth"
+	"github.com/rivo/uniseg"
+)
+
+// State represents the state of the ANSI escape sequence parser used by
+// [DecodeSequence].
+type State = byte
+
+// ANSI escape sequence states used by [DecodeSequence].
+const (
+	NormalState State = iota
+	PrefixState
+	ParamsState
+	IntermedState
+	EscapeState
+	StringState
+)
+
+// DecodeSequence decodes the first ANSI escape sequence or a printable
+// grapheme from the given data. It returns the sequence slice, the number of
+// bytes read, the cell width for each sequence, and the new state.
+//
+// The cell width will always be 0 for control and escape sequences, 1 for
+// ASCII printable characters, and the number of cells other Unicode characters
+// occupy. It uses the uniseg package to calculate the width of Unicode
+// graphemes and characters. This means it will always do grapheme clustering
+// (mode 2027).
+//
+// Passing a non-nil [*Parser] as the last argument will allow the decoder to
+// collect sequence parameters, data, and commands. The parser cmd will have
+// the packed command value that contains intermediate and prefix characters.
+// In the case of a OSC sequence, the cmd will be the OSC command number. Use
+// [Cmd] and [Param] types to unpack command intermediates and prefixes as well
+// as parameters.
+//
+// Zero [Cmd] means the CSI, DCS, or ESC sequence is invalid. Moreover, checking the
+// validity of other data sequences, OSC, DCS, etc, will require checking for
+// the returned sequence terminator bytes such as ST (ESC \\) and BEL).
+//
+// We store the command byte in [Cmd] in the most significant byte, the
+// prefix byte in the next byte, and the intermediate byte in the least
+// significant byte. This is done to avoid using a struct to store the command
+// and its intermediates and prefixes. The command byte is always the least
+// significant byte i.e. [Cmd & 0xff]. Use the [Cmd] type to unpack the
+// command, intermediate, and prefix bytes. Note that we only collect the last
+// prefix character and intermediate byte.
+//
+// The [p.Params] slice will contain the parameters of the sequence. Any
+// sub-parameter will have the [parser.HasMoreFlag] set. Use the [Param] type
+// to unpack the parameters.
+//
+// Example:
+//
+//	var state byte // the initial state is always zero [NormalState]
+//	p := NewParser(32, 1024) // create a new parser with a 32 params buffer and 1024 data buffer (optional)
+//	input := []byte("\x1b[31mHello, World!\x1b[0m")
+//	for len(input) > 0 {
+//		seq, width, n, newState := DecodeSequence(input, state, p)
+//		log.Printf("seq: %q, width: %d", seq, width)
+//		state = newState
+//		input = input[n:]
+//	}
+//
+// This function treats the text as a sequence of grapheme clusters.
+func DecodeSequence[T string | []byte](b T, state byte, p *Parser) (seq T, width int, n int, newState byte) {
+	return decodeSequence(GraphemeWidth, b, state, p)
+}
+
+// DecodeSequenceWc decodes the first ANSI escape sequence or a printable
+// grapheme from the given data. It returns the sequence slice, the number of
+// bytes read, the cell width for each sequence, and the new state.
+//
+// The cell width will always be 0 for control and escape sequences, 1 for
+// ASCII printable characters, and the number of cells other Unicode characters
+// occupy. It uses the uniseg package to calculate the width of Unicode
+// graphemes and characters. This means it will always do grapheme clustering
+// (mode 2027).
+//
+// Passing a non-nil [*Parser] as the last argument will allow the decoder to
+// collect sequence parameters, data, and commands. The parser cmd will have
+// the packed command value that contains intermediate and prefix characters.
+// In the case of a OSC sequence, the cmd will be the OSC command number. Use
+// [Cmd] and [Param] types to unpack command intermediates and prefixes as well
+// as parameters.
+//
+// Zero [Cmd] means the CSI, DCS, or ESC sequence is invalid. Moreover, checking the
+// validity of other data sequences, OSC, DCS, etc, will require checking for
+// the returned sequence terminator bytes such as ST (ESC \\) and BEL).
+//
+// We store the command byte in [Cmd] in the most significant byte, the
+// prefix byte in the next byte, and the intermediate byte in the least
+// significant byte. This is done to avoid using a struct to store the command
+// and its intermediates and prefixes. The command byte is always the least
+// significant byte i.e. [Cmd & 0xff]. Use the [Cmd] type to unpack the
+// command, intermediate, and prefix bytes. Note that we only collect the last
+// prefix character and intermediate byte.
+//
+// The [p.Params] slice will contain the parameters of the sequence. Any
+// sub-parameter will have the [parser.HasMoreFlag] set. Use the [Param] type
+// to unpack the parameters.
+//
+// Example:
+//
+//	var state byte // the initial state is always zero [NormalState]
+//	p := NewParser(32, 1024) // create a new parser with a 32 params buffer and 1024 data buffer (optional)
+//	input := []byte("\x1b[31mHello, World!\x1b[0m")
+//	for len(input) > 0 {
+//		seq, width, n, newState := DecodeSequenceWc(input, state, p)
+//		log.Printf("seq: %q, width: %d", seq, width)
+//		state = newState
+//		input = input[n:]
+//	}
+//
+// This function treats the text as a sequence of wide characters and runes.
+func DecodeSequenceWc[T string | []byte](b T, state byte, p *Parser) (seq T, width int, n int, newState byte) {
+	return decodeSequence(WcWidth, b, state, p)
+}
+
+func decodeSequence[T string | []byte](m Method, b T, state State, p *Parser) (seq T, width int, n int, newState byte) {
+	for i := 0; i < len(b); i++ {
+		c := b[i]
+
+		switch state {
+		case NormalState:
+			switch c {
+			case ESC:
+				if p != nil {
+					if len(p.params) > 0 {
+						p.params[0] = parser.MissingParam
+					}
+					p.cmd = 0
+					p.paramsLen = 0
+					p.dataLen = 0
+				}
+				state = EscapeState
+				continue
+			case CSI, DCS:
+				if p != nil {
+					if len(p.params) > 0 {
+						p.params[0] = parser.MissingParam
+					}
+					p.cmd = 0
+					p.paramsLen = 0
+					p.dataLen = 0
+				}
+				state = PrefixState
+				continue
+			case OSC, APC, SOS, PM:
+				if p != nil {
+					p.cmd = parser.MissingCommand
+					p.dataLen = 0
+				}
+				state = StringState
+				continue
+			}
+
+			if p != nil {
+				p.dataLen = 0
+				p.paramsLen = 0
+				p.cmd = 0
+			}
+			if c > US && c < DEL {
+				// ASCII printable characters
+				return b[i : i+1], 1, 1, NormalState
+			}
+
+			if c <= US || c == DEL || c < 0xC0 {
+				// C0 & C1 control characters & DEL
+				return b[i : i+1], 0, 1, NormalState
+			}
+
+			if utf8.RuneStart(c) {
+				seq, _, width, _ = FirstGraphemeCluster(b, -1)
+				if m == WcWidth {
+					width = runewidth.StringWidth(string(seq))
+				}
+				i += len(seq)
+				return b[:i], width, i, NormalState
+			}
+
+			// Invalid UTF-8 sequence
+			return b[:i], 0, i, NormalState
+		case PrefixState:
+			if c >= '<' && c <= '?' {
+				if p != nil {
+					// We only collect the last prefix character.
+					p.cmd &^= 0xff << parser.PrefixShift
+					p.cmd |= int(c) << parser.PrefixShift
+				}
+				break
+			}
+
+			state = ParamsState
+			fallthrough
+		case ParamsState:
+			if c >= '0' && c <= '9' {
+				if p != nil {
+					if p.params[p.paramsLen] == parser.MissingParam {
+						p.params[p.paramsLen] = 0
+					}
+
+					p.params[p.paramsLen] *= 10
+					p.params[p.paramsLen] += int(c - '0')
+				}
+				break
+			}
+
+			if c == ':' {
+				if p != nil {
+					p.params[p.paramsLen] |= parser.HasMoreFlag
+				}
+			}
+
+			if c == ';' || c == ':' {
+				if p != nil {
+					p.paramsLen++
+					if p.paramsLen < len(p.params) {
+						p.params[p.paramsLen] = parser.MissingParam
+					}
+				}
+				break
+			}
+
+			state = IntermedState
+			fallthrough
+		case IntermedState:
+			if c >= ' ' && c <= '/' {
+				if p != nil {
+					p.cmd &^= 0xff << parser.IntermedShift
+					p.cmd |= int(c) << parser.IntermedShift
+				}
+				break
+			}
+
+			if p != nil {
+				// Increment the last parameter
+				if p.paramsLen > 0 && p.paramsLen < len(p.params)-1 ||
+					p.paramsLen == 0 && len(p.params) > 0 && p.params[0] != parser.MissingParam {
+					p.paramsLen++
+				}
+			}
+
+			if c >= '@' && c <= '~' {
+				if p != nil {
+					p.cmd &^= 0xff
+					p.cmd |= int(c)
+				}
+
+				if HasDcsPrefix(b) {
+					// Continue to collect DCS data
+					if p != nil {
+						p.dataLen = 0
+					}
+					state = StringState
+					continue
+				}
+
+				return b[:i+1], 0, i + 1, NormalState
+			}
+
+			// Invalid CSI/DCS sequence
+			return b[:i], 0, i, NormalState
+		case EscapeState:
+			switch c {
+			case '[', 'P':
+				if p != nil {
+					if len(p.params) > 0 {
+						p.params[0] = parser.MissingParam
+					}
+					p.paramsLen = 0
+					p.cmd = 0
+				}
+				state = PrefixState
+				continue
+			case ']', 'X', '^', '_':
+				if p != nil {
+					p.cmd = parser.MissingCommand
+					p.dataLen = 0
+				}
+				state = StringState
+				continue
+			}
+
+			if c >= ' ' && c <= '/' {
+				if p != nil {
+					p.cmd &^= 0xff << parser.IntermedShift
+					p.cmd |= int(c) << parser.IntermedShift
+				}
+				continue
+			} else if c >= '0' && c <= '~' {
+				if p != nil {
+					p.cmd &^= 0xff
+					p.cmd |= int(c)
+				}
+				return b[:i+1], 0, i + 1, NormalState
+			}
+
+			// Invalid escape sequence
+			return b[:i], 0, i, NormalState
+		case StringState:
+			switch c {
+			case BEL:
+				if HasOscPrefix(b) {
+					parseOscCmd(p)
+					return b[:i+1], 0, i + 1, NormalState
+				}
+			case CAN, SUB:
+				if HasOscPrefix(b) {
+					// Ensure we parse the OSC command number
+					parseOscCmd(p)
+				}
+
+				// Cancel the sequence
+				return b[:i], 0, i, NormalState
+			case ST:
+				if HasOscPrefix(b) {
+					// Ensure we parse the OSC command number
+					parseOscCmd(p)
+				}
+
+				return b[:i+1], 0, i + 1, NormalState
+			case ESC:
+				if HasStPrefix(b[i:]) {
+					if HasOscPrefix(b) {
+						// Ensure we parse the OSC command number
+						parseOscCmd(p)
+					}
+
+					// End of string 7-bit (ST)
+					return b[:i+2], 0, i + 2, NormalState
+				}
+
+				// Otherwise, cancel the sequence
+				return b[:i], 0, i, NormalState
+			}
+
+			if p != nil && p.dataLen < len(p.data) {
+				p.data[p.dataLen] = c
+				p.dataLen++
+
+				// Parse the OSC command number
+				if c == ';' && HasOscPrefix(b) {
+					parseOscCmd(p)
+				}
+			}
+		}
+	}
+
+	return b, 0, len(b), state
+}
+
+func parseOscCmd(p *Parser) {
+	if p == nil || p.cmd != parser.MissingCommand {
+		return
+	}
+	for j := 0; j < p.dataLen; j++ {
+		d := p.data[j]
+		if d < '0' || d > '9' {
+			break
+		}
+		if p.cmd == parser.MissingCommand {
+			p.cmd = 0
+		}
+		p.cmd *= 10
+		p.cmd += int(d - '0')
+	}
+}
+
+// Equal returns true if the given byte slices are equal.
+func Equal[T string | []byte](a, b T) bool {
+	return string(a) == string(b)
+}
+
+// HasPrefix returns true if the given byte slice has prefix.
+func HasPrefix[T string | []byte](b, prefix T) bool {
+	return len(b) >= len(prefix) && Equal(b[0:len(prefix)], prefix)
+}
+
+// HasSuffix returns true if the given byte slice has suffix.
+func HasSuffix[T string | []byte](b, suffix T) bool {
+	return len(b) >= len(suffix) && Equal(b[len(b)-len(suffix):], suffix)
+}
+
+// HasCsiPrefix returns true if the given byte slice has a CSI prefix.
+func HasCsiPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == CSI) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == '[')
+}
+
+// HasOscPrefix returns true if the given byte slice has an OSC prefix.
+func HasOscPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == OSC) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == ']')
+}
+
+// HasApcPrefix returns true if the given byte slice has an APC prefix.
+func HasApcPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == APC) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == '_')
+}
+
+// HasDcsPrefix returns true if the given byte slice has a DCS prefix.
+func HasDcsPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == DCS) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == 'P')
+}
+
+// HasSosPrefix returns true if the given byte slice has a SOS prefix.
+func HasSosPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == SOS) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == 'X')
+}
+
+// HasPmPrefix returns true if the given byte slice has a PM prefix.
+func HasPmPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == PM) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == '^')
+}
+
+// HasStPrefix returns true if the given byte slice has a ST prefix.
+func HasStPrefix[T string | []byte](b T) bool {
+	return (len(b) > 0 && b[0] == ST) ||
+		(len(b) > 1 && b[0] == ESC && b[1] == '\\')
+}
+
+// HasEscPrefix returns true if the given byte slice has an ESC prefix.
+func HasEscPrefix[T string | []byte](b T) bool {
+	return len(b) > 0 && b[0] == ESC
+}
+
+// FirstGraphemeCluster returns the first grapheme cluster in the given string or byte slice.
+// This is a syntactic sugar function that wraps
+// uniseg.FirstGraphemeClusterInString and uniseg.FirstGraphemeCluster.
+func FirstGraphemeCluster[T string | []byte](b T, state int) (T, T, int, int) {
+	switch b := any(b).(type) {
+	case string:
+		cluster, rest, width, newState := uniseg.FirstGraphemeClusterInString(b, state)
+		return T(cluster), T(rest), width, newState
+	case []byte:
+		cluster, rest, width, newState := uniseg.FirstGraphemeCluster(b, state)
+		return T(cluster), T(rest), width, newState
+	}
+	panic("unreachable")
+}
+
+// Cmd represents a sequence command. This is used to pack/unpack a sequence
+// command with its intermediate and prefix characters. Those are commonly
+// found in CSI and DCS sequences.
+type Cmd int
+
+// Prefix returns the unpacked prefix byte of the CSI sequence.
+// This is always gonna be one of the following '<' '=' '>' '?' and in the
+// range of 0x3C-0x3F.
+// Zero is returned if the sequence does not have a prefix.
+func (c Cmd) Prefix() byte {
+	return byte(parser.Prefix(int(c)))
+}
+
+// Intermediate returns the unpacked intermediate byte of the CSI sequence.
+// An intermediate byte is in the range of 0x20-0x2F. This includes these
+// characters from ' ', '!', '"', '#', '$', '%', '&', ”', '(', ')', '*', '+',
+// ',', '-', '.', '/'.
+// Zero is returned if the sequence does not have an intermediate byte.
+func (c Cmd) Intermediate() byte {
+	return byte(parser.Intermediate(int(c)))
+}
+
+// Final returns the unpacked command byte of the CSI sequence.
+func (c Cmd) Final() byte {
+	return byte(parser.Command(int(c)))
+}
+
+// Command packs a command with the given prefix, intermediate, and final. A
+// zero byte means the sequence does not have a prefix or intermediate.
+//
+// Prefixes are in the range of 0x3C-0x3F that is one of `<=>?`.
+//
+// Intermediates are in the range of 0x20-0x2F that is anything in
+// `!"#$%&'()*+,-./`.
+//
+// Final bytes are in the range of 0x40-0x7E that is anything in the range
+// `@A–Z[\]^_`a–z{|}~`.
+func Command(prefix, inter, final byte) (c int) {
+	c = int(final)
+	c |= int(prefix) << parser.PrefixShift
+	c |= int(inter) << parser.IntermedShift
+	return
+}
+
+// Param represents a sequence parameter. Sequence parameters with
+// sub-parameters are packed with the HasMoreFlag set. This is used to unpack
+// the parameters from a CSI and DCS sequences.
+type Param int
+
+// Param returns the unpacked parameter at the given index.
+// It returns the default value if the parameter is missing.
+func (s Param) Param(def int) int {
+	p := int(s) & parser.ParamMask
+	if p == parser.MissingParam {
+		return def
+	}
+	return p
+}
+
+// HasMore unpacks the HasMoreFlag from the parameter.
+func (s Param) HasMore() bool {
+	return s&parser.HasMoreFlag != 0
+}
+
+// Parameter packs an escape code parameter with the given parameter and
+// whether this parameter has following sub-parameters.
+func Parameter(p int, hasMore bool) (s int) {
+	s = p & parser.ParamMask
+	if hasMore {
+		s |= parser.HasMoreFlag
+	}
+	return
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser_handler.go b/vendor/github.com/charmbracelet/x/ansi/parser_handler.go
new file mode 100644
index 0000000..03f9ed4
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser_handler.go
@@ -0,0 +1,60 @@
+package ansi
+
+import "unsafe"
+
+// Params represents a list of packed parameters.
+type Params []Param
+
+// Param returns the parameter at the given index and if it is part of a
+// sub-parameters. It falls back to the default value if the parameter is
+// missing. If the index is out of bounds, it returns the default value and
+// false.
+func (p Params) Param(i, def int) (int, bool, bool) {
+	if i < 0 || i >= len(p) {
+		return def, false, false
+	}
+	return p[i].Param(def), p[i].HasMore(), true
+}
+
+// ForEach iterates over the parameters and calls the given function for each
+// parameter. If a parameter is part of a sub-parameter, it will be called with
+// hasMore set to true.
+// Use def to set a default value for missing parameters.
+func (p Params) ForEach(def int, f func(i, param int, hasMore bool)) {
+	for i := range p {
+		f(i, p[i].Param(def), p[i].HasMore())
+	}
+}
+
+// ToParams converts a list of integers to a list of parameters.
+func ToParams(params []int) Params {
+	return unsafe.Slice((*Param)(unsafe.Pointer(&params[0])), len(params))
+}
+
+// Handler handles actions performed by the parser.
+// It is used to handle ANSI escape sequences, control characters, and runes.
+type Handler struct {
+	// Print is called when a printable rune is encountered.
+	Print func(r rune)
+	// Execute is called when a control character is encountered.
+	Execute func(b byte)
+	// HandleCsi is called when a CSI sequence is encountered.
+	HandleCsi func(cmd Cmd, params Params)
+	// HandleEsc is called when an ESC sequence is encountered.
+	HandleEsc func(cmd Cmd)
+	// HandleDcs is called when a DCS sequence is encountered.
+	HandleDcs func(cmd Cmd, params Params, data []byte)
+	// HandleOsc is called when an OSC sequence is encountered.
+	HandleOsc func(cmd int, data []byte)
+	// HandlePm is called when a PM sequence is encountered.
+	HandlePm func(data []byte)
+	// HandleApc is called when an APC sequence is encountered.
+	HandleApc func(data []byte)
+	// HandleSos is called when a SOS sequence is encountered.
+	HandleSos func(data []byte)
+}
+
+// SetHandler sets the handler for the parser.
+func (p *Parser) SetHandler(h Handler) {
+	p.handler = h
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/parser_sync.go b/vendor/github.com/charmbracelet/x/ansi/parser_sync.go
new file mode 100644
index 0000000..65d25a9
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/parser_sync.go
@@ -0,0 +1,29 @@
+package ansi
+
+import (
+	"sync"
+
+	"github.com/charmbracelet/x/ansi/parser"
+)
+
+var parserPool = sync.Pool{
+	New: func() any {
+		p := NewParser()
+		p.SetParamsSize(parser.MaxParamsSize)
+		p.SetDataSize(1024 * 1024 * 4) // 4MB of data buffer
+		return p
+	},
+}
+
+// GetParser returns a parser from a sync pool.
+func GetParser() *Parser {
+	return parserPool.Get().(*Parser)
+}
+
+// PutParser returns a parser to a sync pool. The parser is reset
+// automatically.
+func PutParser(p *Parser) {
+	p.Reset()
+	p.dataLen = 0
+	parserPool.Put(p)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/passthrough.go b/vendor/github.com/charmbracelet/x/ansi/passthrough.go
new file mode 100644
index 0000000..14a7452
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/passthrough.go
@@ -0,0 +1,63 @@
+package ansi
+
+import (
+	"bytes"
+)
+
+// ScreenPassthrough wraps the given ANSI sequence in a DCS passthrough
+// sequence to be sent to the outer terminal. This is used to send raw escape
+// sequences to the outer terminal when running inside GNU Screen.
+//
+//	DCS <data> ST
+//
+// Note: Screen limits the length of string sequences to 768 bytes (since 2014).
+// Use zero to indicate no limit, otherwise, this will chunk the returned
+// string into limit sized chunks.
+//
+// See: https://www.gnu.org/software/screen/manual/screen.html#String-Escapes
+// See: https://git.savannah.gnu.org/cgit/screen.git/tree/src/screen.h?id=c184c6ec27683ff1a860c45be5cf520d896fd2ef#n44
+func ScreenPassthrough(seq string, limit int) string {
+	var b bytes.Buffer
+	b.WriteString("\x1bP")
+	if limit > 0 {
+		for i := 0; i < len(seq); i += limit {
+			end := i + limit
+			if end > len(seq) {
+				end = len(seq)
+			}
+			b.WriteString(seq[i:end])
+			if end < len(seq) {
+				b.WriteString("\x1b\\\x1bP")
+			}
+		}
+	} else {
+		b.WriteString(seq)
+	}
+	b.WriteString("\x1b\\")
+	return b.String()
+}
+
+// TmuxPassthrough wraps the given ANSI sequence in a special DCS passthrough
+// sequence to be sent to the outer terminal. This is used to send raw escape
+// sequences to the outer terminal when running inside Tmux.
+//
+//	DCS tmux ; <escaped-data> ST
+//
+// Where <escaped-data> is the given sequence in which all occurrences of ESC
+// (0x1b) are doubled i.e. replaced with ESC ESC (0x1b 0x1b).
+//
+// Note: this needs the `allow-passthrough` option to be set to `on`.
+//
+// See: https://github.com/tmux/tmux/wiki/FAQ#what-is-the-passthrough-escape-sequence-and-how-do-i-use-it
+func TmuxPassthrough(seq string) string {
+	var b bytes.Buffer
+	b.WriteString("\x1bPtmux;")
+	for i := 0; i < len(seq); i++ {
+		if seq[i] == ESC {
+			b.WriteByte(ESC)
+		}
+		b.WriteByte(seq[i])
+	}
+	b.WriteString("\x1b\\")
+	return b.String()
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/paste.go b/vendor/github.com/charmbracelet/x/ansi/paste.go
new file mode 100644
index 0000000..2f9ea6f
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/paste.go
@@ -0,0 +1,7 @@
+package ansi
+
+// BracketedPasteStart is the control sequence to enable bracketed paste mode.
+const BracketedPasteStart = "\x1b[200~"
+
+// BracketedPasteEnd is the control sequence to disable bracketed paste mode.
+const BracketedPasteEnd = "\x1b[201~"
diff --git a/vendor/github.com/charmbracelet/x/ansi/reset.go b/vendor/github.com/charmbracelet/x/ansi/reset.go
new file mode 100644
index 0000000..c1b89ea
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/reset.go
@@ -0,0 +1,11 @@
+package ansi
+
+// ResetInitialState (RIS) resets the terminal to its initial state.
+//
+//	ESC c
+//
+// See: https://vt100.net/docs/vt510-rm/RIS.html
+const (
+	ResetInitialState = "\x1bc"
+	RIS               = ResetInitialState
+)
diff --git a/vendor/github.com/charmbracelet/x/ansi/screen.go b/vendor/github.com/charmbracelet/x/ansi/screen.go
new file mode 100644
index 0000000..c76e4f0
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/screen.go
@@ -0,0 +1,410 @@
+package ansi
+
+import (
+	"strconv"
+	"strings"
+)
+
+// EraseDisplay (ED) clears the display or parts of the display. A screen is
+// the shown part of the terminal display excluding the scrollback buffer.
+// Possible values:
+//
+// Default is 0.
+//
+//	 0: Clear from cursor to end of screen.
+//	 1: Clear from cursor to beginning of the screen.
+//	 2: Clear entire screen (and moves cursor to upper left on DOS).
+//	 3: Clear entire display which delete all lines saved in the scrollback buffer (xterm).
+//
+//	CSI <n> J
+//
+// See: https://vt100.net/docs/vt510-rm/ED.html
+func EraseDisplay(n int) string {
+	var s string
+	if n > 0 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "J"
+}
+
+// ED is an alias for [EraseDisplay].
+func ED(n int) string {
+	return EraseDisplay(n)
+}
+
+// EraseDisplay constants.
+// These are the possible values for the EraseDisplay function.
+const (
+	EraseScreenBelow   = "\x1b[J"
+	EraseScreenAbove   = "\x1b[1J"
+	EraseEntireScreen  = "\x1b[2J"
+	EraseEntireDisplay = "\x1b[3J"
+)
+
+// EraseLine (EL) clears the current line or parts of the line. Possible values:
+//
+//	0: Clear from cursor to end of line.
+//	1: Clear from cursor to beginning of the line.
+//	2: Clear entire line.
+//
+// The cursor position is not affected.
+//
+//	CSI <n> K
+//
+// See: https://vt100.net/docs/vt510-rm/EL.html
+func EraseLine(n int) string {
+	var s string
+	if n > 0 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "K"
+}
+
+// EL is an alias for [EraseLine].
+func EL(n int) string {
+	return EraseLine(n)
+}
+
+// EraseLine constants.
+// These are the possible values for the EraseLine function.
+const (
+	EraseLineRight  = "\x1b[K"
+	EraseLineLeft   = "\x1b[1K"
+	EraseEntireLine = "\x1b[2K"
+)
+
+// ScrollUp (SU) scrolls the screen up n lines. New lines are added at the
+// bottom of the screen.
+//
+//	CSI Pn S
+//
+// See: https://vt100.net/docs/vt510-rm/SU.html
+func ScrollUp(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "S"
+}
+
+// PanDown is an alias for [ScrollUp].
+func PanDown(n int) string {
+	return ScrollUp(n)
+}
+
+// SU is an alias for [ScrollUp].
+func SU(n int) string {
+	return ScrollUp(n)
+}
+
+// ScrollDown (SD) scrolls the screen down n lines. New lines are added at the
+// top of the screen.
+//
+//	CSI Pn T
+//
+// See: https://vt100.net/docs/vt510-rm/SD.html
+func ScrollDown(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "T"
+}
+
+// PanUp is an alias for [ScrollDown].
+func PanUp(n int) string {
+	return ScrollDown(n)
+}
+
+// SD is an alias for [ScrollDown].
+func SD(n int) string {
+	return ScrollDown(n)
+}
+
+// InsertLine (IL) inserts n blank lines at the current cursor position.
+// Existing lines are moved down.
+//
+//	CSI Pn L
+//
+// See: https://vt100.net/docs/vt510-rm/IL.html
+func InsertLine(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "L"
+}
+
+// IL is an alias for [InsertLine].
+func IL(n int) string {
+	return InsertLine(n)
+}
+
+// DeleteLine (DL) deletes n lines at the current cursor position. Existing
+// lines are moved up.
+//
+//	CSI Pn M
+//
+// See: https://vt100.net/docs/vt510-rm/DL.html
+func DeleteLine(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "M"
+}
+
+// DL is an alias for [DeleteLine].
+func DL(n int) string {
+	return DeleteLine(n)
+}
+
+// SetTopBottomMargins (DECSTBM) sets the top and bottom margins for the scrolling
+// region. The default is the entire screen.
+//
+// Default is 1 and the bottom of the screen.
+//
+//	CSI Pt ; Pb r
+//
+// See: https://vt100.net/docs/vt510-rm/DECSTBM.html
+func SetTopBottomMargins(top, bot int) string {
+	var t, b string
+	if top > 0 {
+		t = strconv.Itoa(top)
+	}
+	if bot > 0 {
+		b = strconv.Itoa(bot)
+	}
+	return "\x1b[" + t + ";" + b + "r"
+}
+
+// DECSTBM is an alias for [SetTopBottomMargins].
+func DECSTBM(top, bot int) string {
+	return SetTopBottomMargins(top, bot)
+}
+
+// SetLeftRightMargins (DECSLRM) sets the left and right margins for the scrolling
+// region.
+//
+// Default is 1 and the right of the screen.
+//
+//	CSI Pl ; Pr s
+//
+// See: https://vt100.net/docs/vt510-rm/DECSLRM.html
+func SetLeftRightMargins(left, right int) string {
+	var l, r string
+	if left > 0 {
+		l = strconv.Itoa(left)
+	}
+	if right > 0 {
+		r = strconv.Itoa(right)
+	}
+	return "\x1b[" + l + ";" + r + "s"
+}
+
+// DECSLRM is an alias for [SetLeftRightMargins].
+func DECSLRM(left, right int) string {
+	return SetLeftRightMargins(left, right)
+}
+
+// SetScrollingRegion (DECSTBM) sets the top and bottom margins for the scrolling
+// region. The default is the entire screen.
+//
+//	CSI <top> ; <bottom> r
+//
+// See: https://vt100.net/docs/vt510-rm/DECSTBM.html
+//
+// Deprecated: use [SetTopBottomMargins] instead.
+func SetScrollingRegion(t, b int) string {
+	if t < 0 {
+		t = 0
+	}
+	if b < 0 {
+		b = 0
+	}
+	return "\x1b[" + strconv.Itoa(t) + ";" + strconv.Itoa(b) + "r"
+}
+
+// InsertCharacter (ICH) inserts n blank characters at the current cursor
+// position. Existing characters move to the right. Characters moved past the
+// right margin are lost. ICH has no effect outside the scrolling margins.
+//
+// Default is 1.
+//
+//	CSI Pn @
+//
+// See: https://vt100.net/docs/vt510-rm/ICH.html
+func InsertCharacter(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "@"
+}
+
+// ICH is an alias for [InsertCharacter].
+func ICH(n int) string {
+	return InsertCharacter(n)
+}
+
+// DeleteCharacter (DCH) deletes n characters at the current cursor position.
+// As the characters are deleted, the remaining characters move to the left and
+// the cursor remains at the same position.
+//
+// Default is 1.
+//
+//	CSI Pn P
+//
+// See: https://vt100.net/docs/vt510-rm/DCH.html
+func DeleteCharacter(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "P"
+}
+
+// DCH is an alias for [DeleteCharacter].
+func DCH(n int) string {
+	return DeleteCharacter(n)
+}
+
+// SetTabEvery8Columns (DECST8C) sets the tab stops at every 8 columns.
+//
+//	CSI ? 5 W
+//
+// See: https://vt100.net/docs/vt510-rm/DECST8C.html
+const (
+	SetTabEvery8Columns = "\x1b[?5W"
+	DECST8C             = SetTabEvery8Columns
+)
+
+// HorizontalTabSet (HTS) sets a horizontal tab stop at the current cursor
+// column.
+//
+// This is equivalent to [HTS].
+//
+//	ESC H
+//
+// See: https://vt100.net/docs/vt510-rm/HTS.html
+const HorizontalTabSet = "\x1bH"
+
+// TabClear (TBC) clears tab stops.
+//
+// Default is 0.
+//
+// Possible values:
+// 0: Clear tab stop at the current column. (default)
+// 3: Clear all tab stops.
+//
+//	CSI Pn g
+//
+// See: https://vt100.net/docs/vt510-rm/TBC.html
+func TabClear(n int) string {
+	var s string
+	if n > 0 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "g"
+}
+
+// TBC is an alias for [TabClear].
+func TBC(n int) string {
+	return TabClear(n)
+}
+
+// RequestPresentationStateReport (DECRQPSR) requests the terminal to send a
+// report of the presentation state. This includes the cursor information [DECCIR],
+// and tab stop [DECTABSR] reports.
+//
+// Default is 0.
+//
+// Possible values:
+// 0: Error, request ignored.
+// 1: Cursor information report [DECCIR].
+// 2: Tab stop report [DECTABSR].
+//
+//	CSI Ps $ w
+//
+// See: https://vt100.net/docs/vt510-rm/DECRQPSR.html
+func RequestPresentationStateReport(n int) string {
+	var s string
+	if n > 0 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "$w"
+}
+
+// DECRQPSR is an alias for [RequestPresentationStateReport].
+func DECRQPSR(n int) string {
+	return RequestPresentationStateReport(n)
+}
+
+// TabStopReport (DECTABSR) is the response to a tab stop report request.
+// It reports the tab stops set in the terminal.
+//
+// The response is a list of tab stops separated by a slash (/) character.
+//
+//	DCS 2 $ u D ... D ST
+//
+// Where D is a decimal number representing a tab stop.
+//
+// See: https://vt100.net/docs/vt510-rm/DECTABSR.html
+func TabStopReport(stops ...int) string {
+	var s []string
+	for _, v := range stops {
+		s = append(s, strconv.Itoa(v))
+	}
+	return "\x1bP2$u" + strings.Join(s, "/") + "\x1b\\"
+}
+
+// DECTABSR is an alias for [TabStopReport].
+func DECTABSR(stops ...int) string {
+	return TabStopReport(stops...)
+}
+
+// CursorInformationReport (DECCIR) is the response to a cursor information
+// report request. It reports the cursor position, visual attributes, and
+// character protection attributes. It also reports the status of origin mode
+// [DECOM] and the current active character set.
+//
+// The response is a list of values separated by a semicolon (;) character.
+//
+//	DCS 1 $ u D ... D ST
+//
+// Where D is a decimal number representing a value.
+//
+// See: https://vt100.net/docs/vt510-rm/DECCIR.html
+func CursorInformationReport(values ...int) string {
+	var s []string
+	for _, v := range values {
+		s = append(s, strconv.Itoa(v))
+	}
+	return "\x1bP1$u" + strings.Join(s, ";") + "\x1b\\"
+}
+
+// DECCIR is an alias for [CursorInformationReport].
+func DECCIR(values ...int) string {
+	return CursorInformationReport(values...)
+}
+
+// RepeatPreviousCharacter (REP) repeats the previous character n times.
+// This is identical to typing the same character n times.
+//
+// Default is 1.
+//
+//	CSI Pn b
+//
+// See: ECMA-48 § 8.3.103
+func RepeatPreviousCharacter(n int) string {
+	var s string
+	if n > 1 {
+		s = strconv.Itoa(n)
+	}
+	return "\x1b[" + s + "b"
+}
+
+// REP is an alias for [RepeatPreviousCharacter].
+func REP(n int) string {
+	return RepeatPreviousCharacter(n)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/sgr.go b/vendor/github.com/charmbracelet/x/ansi/sgr.go
new file mode 100644
index 0000000..1a18c98
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/sgr.go
@@ -0,0 +1,95 @@
+package ansi
+
+import "strconv"
+
+// Select Graphic Rendition (SGR) is a command that sets display attributes.
+//
+// Default is 0.
+//
+//	CSI Ps ; Ps ... m
+//
+// See: https://vt100.net/docs/vt510-rm/SGR.html
+func SelectGraphicRendition(ps ...Attr) string {
+	if len(ps) == 0 {
+		return ResetStyle
+	}
+
+	var s Style
+	for _, p := range ps {
+		attr, ok := attrStrings[p]
+		if ok {
+			s = append(s, attr)
+		} else {
+			if p < 0 {
+				p = 0
+			}
+			s = append(s, strconv.Itoa(p))
+		}
+	}
+
+	return s.String()
+}
+
+// SGR is an alias for [SelectGraphicRendition].
+func SGR(ps ...Attr) string {
+	return SelectGraphicRendition(ps...)
+}
+
+var attrStrings = map[int]string{
+	ResetAttr:                        "0",
+	BoldAttr:                         "1",
+	FaintAttr:                        "2",
+	ItalicAttr:                       "3",
+	UnderlineAttr:                    "4",
+	SlowBlinkAttr:                    "5",
+	RapidBlinkAttr:                   "6",
+	ReverseAttr:                      "7",
+	ConcealAttr:                      "8",
+	StrikethroughAttr:                "9",
+	NoBoldAttr:                       "21",
+	NormalIntensityAttr:              "22",
+	NoItalicAttr:                     "23",
+	NoUnderlineAttr:                  "24",
+	NoBlinkAttr:                      "25",
+	NoReverseAttr:                    "27",
+	NoConcealAttr:                    "28",
+	NoStrikethroughAttr:              "29",
+	BlackForegroundColorAttr:         "30",
+	RedForegroundColorAttr:           "31",
+	GreenForegroundColorAttr:         "32",
+	YellowForegroundColorAttr:        "33",
+	BlueForegroundColorAttr:          "34",
+	MagentaForegroundColorAttr:       "35",
+	CyanForegroundColorAttr:          "36",
+	WhiteForegroundColorAttr:         "37",
+	ExtendedForegroundColorAttr:      "38",
+	DefaultForegroundColorAttr:       "39",
+	BlackBackgroundColorAttr:         "40",
+	RedBackgroundColorAttr:           "41",
+	GreenBackgroundColorAttr:         "42",
+	YellowBackgroundColorAttr:        "43",
+	BlueBackgroundColorAttr:          "44",
+	MagentaBackgroundColorAttr:       "45",
+	CyanBackgroundColorAttr:          "46",
+	WhiteBackgroundColorAttr:         "47",
+	ExtendedBackgroundColorAttr:      "48",
+	DefaultBackgroundColorAttr:       "49",
+	ExtendedUnderlineColorAttr:       "58",
+	DefaultUnderlineColorAttr:        "59",
+	BrightBlackForegroundColorAttr:   "90",
+	BrightRedForegroundColorAttr:     "91",
+	BrightGreenForegroundColorAttr:   "92",
+	BrightYellowForegroundColorAttr:  "93",
+	BrightBlueForegroundColorAttr:    "94",
+	BrightMagentaForegroundColorAttr: "95",
+	BrightCyanForegroundColorAttr:    "96",
+	BrightWhiteForegroundColorAttr:   "97",
+	BrightBlackBackgroundColorAttr:   "100",
+	BrightRedBackgroundColorAttr:     "101",
+	BrightGreenBackgroundColorAttr:   "102",
+	BrightYellowBackgroundColorAttr:  "103",
+	BrightBlueBackgroundColorAttr:    "104",
+	BrightMagentaBackgroundColorAttr: "105",
+	BrightCyanBackgroundColorAttr:    "106",
+	BrightWhiteBackgroundColorAttr:   "107",
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/status.go b/vendor/github.com/charmbracelet/x/ansi/status.go
new file mode 100644
index 0000000..4337e18
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/status.go
@@ -0,0 +1,144 @@
+package ansi
+
+import (
+	"strconv"
+	"strings"
+)
+
+// StatusReport represents a terminal status report.
+type StatusReport interface {
+	// StatusReport returns the status report identifier.
+	StatusReport() int
+}
+
+// ANSIReport represents an ANSI terminal status report.
+type ANSIStatusReport int //nolint:revive
+
+// Report returns the status report identifier.
+func (s ANSIStatusReport) StatusReport() int {
+	return int(s)
+}
+
+// DECStatusReport represents a DEC terminal status report.
+type DECStatusReport int
+
+// Status returns the status report identifier.
+func (s DECStatusReport) StatusReport() int {
+	return int(s)
+}
+
+// DeviceStatusReport (DSR) is a control sequence that reports the terminal's
+// status.
+// The terminal responds with a DSR sequence.
+//
+//	CSI Ps n
+//	CSI ? Ps n
+//
+// If one of the statuses is a [DECStatus], the sequence will use the DEC
+// format.
+//
+// See also https://vt100.net/docs/vt510-rm/DSR.html
+func DeviceStatusReport(statues ...StatusReport) string {
+	var dec bool
+	list := make([]string, len(statues))
+	seq := "\x1b["
+	for i, status := range statues {
+		list[i] = strconv.Itoa(status.StatusReport())
+		switch status.(type) {
+		case DECStatusReport:
+			dec = true
+		}
+	}
+	if dec {
+		seq += "?"
+	}
+	return seq + strings.Join(list, ";") + "n"
+}
+
+// DSR is an alias for [DeviceStatusReport].
+func DSR(status StatusReport) string {
+	return DeviceStatusReport(status)
+}
+
+// RequestCursorPositionReport is an escape sequence that requests the current
+// cursor position.
+//
+//	CSI 6 n
+//
+// The terminal will report the cursor position as a CSI sequence in the
+// following format:
+//
+//	CSI Pl ; Pc R
+//
+// Where Pl is the line number and Pc is the column number.
+// See: https://vt100.net/docs/vt510-rm/CPR.html
+const RequestCursorPositionReport = "\x1b[6n"
+
+// RequestExtendedCursorPositionReport (DECXCPR) is a sequence for requesting
+// the cursor position report including the current page number.
+//
+//	CSI ? 6 n
+//
+// The terminal will report the cursor position as a CSI sequence in the
+// following format:
+//
+//	CSI ? Pl ; Pc ; Pp R
+//
+// Where Pl is the line number, Pc is the column number, and Pp is the page
+// number.
+// See: https://vt100.net/docs/vt510-rm/DECXCPR.html
+const RequestExtendedCursorPositionReport = "\x1b[?6n"
+
+// CursorPositionReport (CPR) is a control sequence that reports the cursor's
+// position.
+//
+//	CSI Pl ; Pc R
+//
+// Where Pl is the line number and Pc is the column number.
+//
+// See also https://vt100.net/docs/vt510-rm/CPR.html
+func CursorPositionReport(line, column int) string {
+	if line < 1 {
+		line = 1
+	}
+	if column < 1 {
+		column = 1
+	}
+	return "\x1b[" + strconv.Itoa(line) + ";" + strconv.Itoa(column) + "R"
+}
+
+// CPR is an alias for [CursorPositionReport].
+func CPR(line, column int) string {
+	return CursorPositionReport(line, column)
+}
+
+// ExtendedCursorPositionReport (DECXCPR) is a control sequence that reports the
+// cursor's position along with the page number (optional).
+//
+//	CSI ? Pl ; Pc R
+//	CSI ? Pl ; Pc ; Pv R
+//
+// Where Pl is the line number, Pc is the column number, and Pv is the page
+// number.
+//
+// If the page number is zero or negative, the returned sequence won't include
+// the page number.
+//
+// See also https://vt100.net/docs/vt510-rm/DECXCPR.html
+func ExtendedCursorPositionReport(line, column, page int) string {
+	if line < 1 {
+		line = 1
+	}
+	if column < 1 {
+		column = 1
+	}
+	if page < 1 {
+		return "\x1b[?" + strconv.Itoa(line) + ";" + strconv.Itoa(column) + "R"
+	}
+	return "\x1b[?" + strconv.Itoa(line) + ";" + strconv.Itoa(column) + ";" + strconv.Itoa(page) + "R"
+}
+
+// DECXCPR is an alias for [ExtendedCursorPositionReport].
+func DECXCPR(line, column, page int) string {
+	return ExtendedCursorPositionReport(line, column, page)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/style.go b/vendor/github.com/charmbracelet/x/ansi/style.go
new file mode 100644
index 0000000..46ddcaa
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/style.go
@@ -0,0 +1,660 @@
+package ansi
+
+import (
+	"image/color"
+	"strconv"
+	"strings"
+)
+
+// ResetStyle is a SGR (Select Graphic Rendition) style sequence that resets
+// all attributes.
+// See: https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_(Select_Graphic_Rendition)_parameters
+const ResetStyle = "\x1b[m"
+
+// Attr is a SGR (Select Graphic Rendition) style attribute.
+type Attr = int
+
+// Style represents an ANSI SGR (Select Graphic Rendition) style.
+type Style []string
+
+// String returns the ANSI SGR (Select Graphic Rendition) style sequence for
+// the given style.
+func (s Style) String() string {
+	if len(s) == 0 {
+		return ResetStyle
+	}
+	return "\x1b[" + strings.Join(s, ";") + "m"
+}
+
+// Styled returns a styled string with the given style applied.
+func (s Style) Styled(str string) string {
+	if len(s) == 0 {
+		return str
+	}
+	return s.String() + str + ResetStyle
+}
+
+// Reset appends the reset style attribute to the style.
+func (s Style) Reset() Style {
+	return append(s, resetAttr)
+}
+
+// Bold appends the bold style attribute to the style.
+func (s Style) Bold() Style {
+	return append(s, boldAttr)
+}
+
+// Faint appends the faint style attribute to the style.
+func (s Style) Faint() Style {
+	return append(s, faintAttr)
+}
+
+// Italic appends the italic style attribute to the style.
+func (s Style) Italic() Style {
+	return append(s, italicAttr)
+}
+
+// Underline appends the underline style attribute to the style.
+func (s Style) Underline() Style {
+	return append(s, underlineAttr)
+}
+
+// UnderlineStyle appends the underline style attribute to the style.
+func (s Style) UnderlineStyle(u UnderlineStyle) Style {
+	switch u {
+	case NoUnderlineStyle:
+		return s.NoUnderline()
+	case SingleUnderlineStyle:
+		return s.Underline()
+	case DoubleUnderlineStyle:
+		return append(s, doubleUnderlineStyle)
+	case CurlyUnderlineStyle:
+		return append(s, curlyUnderlineStyle)
+	case DottedUnderlineStyle:
+		return append(s, dottedUnderlineStyle)
+	case DashedUnderlineStyle:
+		return append(s, dashedUnderlineStyle)
+	}
+	return s
+}
+
+// DoubleUnderline appends the double underline style attribute to the style.
+// This is a convenience method for UnderlineStyle(DoubleUnderlineStyle).
+func (s Style) DoubleUnderline() Style {
+	return s.UnderlineStyle(DoubleUnderlineStyle)
+}
+
+// CurlyUnderline appends the curly underline style attribute to the style.
+// This is a convenience method for UnderlineStyle(CurlyUnderlineStyle).
+func (s Style) CurlyUnderline() Style {
+	return s.UnderlineStyle(CurlyUnderlineStyle)
+}
+
+// DottedUnderline appends the dotted underline style attribute to the style.
+// This is a convenience method for UnderlineStyle(DottedUnderlineStyle).
+func (s Style) DottedUnderline() Style {
+	return s.UnderlineStyle(DottedUnderlineStyle)
+}
+
+// DashedUnderline appends the dashed underline style attribute to the style.
+// This is a convenience method for UnderlineStyle(DashedUnderlineStyle).
+func (s Style) DashedUnderline() Style {
+	return s.UnderlineStyle(DashedUnderlineStyle)
+}
+
+// SlowBlink appends the slow blink style attribute to the style.
+func (s Style) SlowBlink() Style {
+	return append(s, slowBlinkAttr)
+}
+
+// RapidBlink appends the rapid blink style attribute to the style.
+func (s Style) RapidBlink() Style {
+	return append(s, rapidBlinkAttr)
+}
+
+// Reverse appends the reverse style attribute to the style.
+func (s Style) Reverse() Style {
+	return append(s, reverseAttr)
+}
+
+// Conceal appends the conceal style attribute to the style.
+func (s Style) Conceal() Style {
+	return append(s, concealAttr)
+}
+
+// Strikethrough appends the strikethrough style attribute to the style.
+func (s Style) Strikethrough() Style {
+	return append(s, strikethroughAttr)
+}
+
+// NoBold appends the no bold style attribute to the style.
+func (s Style) NoBold() Style {
+	return append(s, noBoldAttr)
+}
+
+// NormalIntensity appends the normal intensity style attribute to the style.
+func (s Style) NormalIntensity() Style {
+	return append(s, normalIntensityAttr)
+}
+
+// NoItalic appends the no italic style attribute to the style.
+func (s Style) NoItalic() Style {
+	return append(s, noItalicAttr)
+}
+
+// NoUnderline appends the no underline style attribute to the style.
+func (s Style) NoUnderline() Style {
+	return append(s, noUnderlineAttr)
+}
+
+// NoBlink appends the no blink style attribute to the style.
+func (s Style) NoBlink() Style {
+	return append(s, noBlinkAttr)
+}
+
+// NoReverse appends the no reverse style attribute to the style.
+func (s Style) NoReverse() Style {
+	return append(s, noReverseAttr)
+}
+
+// NoConceal appends the no conceal style attribute to the style.
+func (s Style) NoConceal() Style {
+	return append(s, noConcealAttr)
+}
+
+// NoStrikethrough appends the no strikethrough style attribute to the style.
+func (s Style) NoStrikethrough() Style {
+	return append(s, noStrikethroughAttr)
+}
+
+// DefaultForegroundColor appends the default foreground color style attribute to the style.
+func (s Style) DefaultForegroundColor() Style {
+	return append(s, defaultForegroundColorAttr)
+}
+
+// DefaultBackgroundColor appends the default background color style attribute to the style.
+func (s Style) DefaultBackgroundColor() Style {
+	return append(s, defaultBackgroundColorAttr)
+}
+
+// DefaultUnderlineColor appends the default underline color style attribute to the style.
+func (s Style) DefaultUnderlineColor() Style {
+	return append(s, defaultUnderlineColorAttr)
+}
+
+// ForegroundColor appends the foreground color style attribute to the style.
+func (s Style) ForegroundColor(c Color) Style {
+	return append(s, foregroundColorString(c))
+}
+
+// BackgroundColor appends the background color style attribute to the style.
+func (s Style) BackgroundColor(c Color) Style {
+	return append(s, backgroundColorString(c))
+}
+
+// UnderlineColor appends the underline color style attribute to the style.
+func (s Style) UnderlineColor(c Color) Style {
+	return append(s, underlineColorString(c))
+}
+
+// UnderlineStyle represents an ANSI SGR (Select Graphic Rendition) underline
+// style.
+type UnderlineStyle = byte
+
+const (
+	doubleUnderlineStyle = "4:2"
+	curlyUnderlineStyle  = "4:3"
+	dottedUnderlineStyle = "4:4"
+	dashedUnderlineStyle = "4:5"
+)
+
+const (
+	// NoUnderlineStyle is the default underline style.
+	NoUnderlineStyle UnderlineStyle = iota
+	// SingleUnderlineStyle is a single underline style.
+	SingleUnderlineStyle
+	// DoubleUnderlineStyle is a double underline style.
+	DoubleUnderlineStyle
+	// CurlyUnderlineStyle is a curly underline style.
+	CurlyUnderlineStyle
+	// DottedUnderlineStyle is a dotted underline style.
+	DottedUnderlineStyle
+	// DashedUnderlineStyle is a dashed underline style.
+	DashedUnderlineStyle
+)
+
+// SGR (Select Graphic Rendition) style attributes.
+// See: https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_(Select_Graphic_Rendition)_parameters
+const (
+	ResetAttr                        Attr = 0
+	BoldAttr                         Attr = 1
+	FaintAttr                        Attr = 2
+	ItalicAttr                       Attr = 3
+	UnderlineAttr                    Attr = 4
+	SlowBlinkAttr                    Attr = 5
+	RapidBlinkAttr                   Attr = 6
+	ReverseAttr                      Attr = 7
+	ConcealAttr                      Attr = 8
+	StrikethroughAttr                Attr = 9
+	NoBoldAttr                       Attr = 21 // Some terminals treat this as double underline.
+	NormalIntensityAttr              Attr = 22
+	NoItalicAttr                     Attr = 23
+	NoUnderlineAttr                  Attr = 24
+	NoBlinkAttr                      Attr = 25
+	NoReverseAttr                    Attr = 27
+	NoConcealAttr                    Attr = 28
+	NoStrikethroughAttr              Attr = 29
+	BlackForegroundColorAttr         Attr = 30
+	RedForegroundColorAttr           Attr = 31
+	GreenForegroundColorAttr         Attr = 32
+	YellowForegroundColorAttr        Attr = 33
+	BlueForegroundColorAttr          Attr = 34
+	MagentaForegroundColorAttr       Attr = 35
+	CyanForegroundColorAttr          Attr = 36
+	WhiteForegroundColorAttr         Attr = 37
+	ExtendedForegroundColorAttr      Attr = 38
+	DefaultForegroundColorAttr       Attr = 39
+	BlackBackgroundColorAttr         Attr = 40
+	RedBackgroundColorAttr           Attr = 41
+	GreenBackgroundColorAttr         Attr = 42
+	YellowBackgroundColorAttr        Attr = 43
+	BlueBackgroundColorAttr          Attr = 44
+	MagentaBackgroundColorAttr       Attr = 45
+	CyanBackgroundColorAttr          Attr = 46
+	WhiteBackgroundColorAttr         Attr = 47
+	ExtendedBackgroundColorAttr      Attr = 48
+	DefaultBackgroundColorAttr       Attr = 49
+	ExtendedUnderlineColorAttr       Attr = 58
+	DefaultUnderlineColorAttr        Attr = 59
+	BrightBlackForegroundColorAttr   Attr = 90
+	BrightRedForegroundColorAttr     Attr = 91
+	BrightGreenForegroundColorAttr   Attr = 92
+	BrightYellowForegroundColorAttr  Attr = 93
+	BrightBlueForegroundColorAttr    Attr = 94
+	BrightMagentaForegroundColorAttr Attr = 95
+	BrightCyanForegroundColorAttr    Attr = 96
+	BrightWhiteForegroundColorAttr   Attr = 97
+	BrightBlackBackgroundColorAttr   Attr = 100
+	BrightRedBackgroundColorAttr     Attr = 101
+	BrightGreenBackgroundColorAttr   Attr = 102
+	BrightYellowBackgroundColorAttr  Attr = 103
+	BrightBlueBackgroundColorAttr    Attr = 104
+	BrightMagentaBackgroundColorAttr Attr = 105
+	BrightCyanBackgroundColorAttr    Attr = 106
+	BrightWhiteBackgroundColorAttr   Attr = 107
+
+	RGBColorIntroducerAttr      Attr = 2
+	ExtendedColorIntroducerAttr Attr = 5
+)
+
+const (
+	resetAttr                        = "0"
+	boldAttr                         = "1"
+	faintAttr                        = "2"
+	italicAttr                       = "3"
+	underlineAttr                    = "4"
+	slowBlinkAttr                    = "5"
+	rapidBlinkAttr                   = "6"
+	reverseAttr                      = "7"
+	concealAttr                      = "8"
+	strikethroughAttr                = "9"
+	noBoldAttr                       = "21"
+	normalIntensityAttr              = "22"
+	noItalicAttr                     = "23"
+	noUnderlineAttr                  = "24"
+	noBlinkAttr                      = "25"
+	noReverseAttr                    = "27"
+	noConcealAttr                    = "28"
+	noStrikethroughAttr              = "29"
+	blackForegroundColorAttr         = "30"
+	redForegroundColorAttr           = "31"
+	greenForegroundColorAttr         = "32"
+	yellowForegroundColorAttr        = "33"
+	blueForegroundColorAttr          = "34"
+	magentaForegroundColorAttr       = "35"
+	cyanForegroundColorAttr          = "36"
+	whiteForegroundColorAttr         = "37"
+	extendedForegroundColorAttr      = "38"
+	defaultForegroundColorAttr       = "39"
+	blackBackgroundColorAttr         = "40"
+	redBackgroundColorAttr           = "41"
+	greenBackgroundColorAttr         = "42"
+	yellowBackgroundColorAttr        = "43"
+	blueBackgroundColorAttr          = "44"
+	magentaBackgroundColorAttr       = "45"
+	cyanBackgroundColorAttr          = "46"
+	whiteBackgroundColorAttr         = "47"
+	extendedBackgroundColorAttr      = "48"
+	defaultBackgroundColorAttr       = "49"
+	extendedUnderlineColorAttr       = "58"
+	defaultUnderlineColorAttr        = "59"
+	brightBlackForegroundColorAttr   = "90"
+	brightRedForegroundColorAttr     = "91"
+	brightGreenForegroundColorAttr   = "92"
+	brightYellowForegroundColorAttr  = "93"
+	brightBlueForegroundColorAttr    = "94"
+	brightMagentaForegroundColorAttr = "95"
+	brightCyanForegroundColorAttr    = "96"
+	brightWhiteForegroundColorAttr   = "97"
+	brightBlackBackgroundColorAttr   = "100"
+	brightRedBackgroundColorAttr     = "101"
+	brightGreenBackgroundColorAttr   = "102"
+	brightYellowBackgroundColorAttr  = "103"
+	brightBlueBackgroundColorAttr    = "104"
+	brightMagentaBackgroundColorAttr = "105"
+	brightCyanBackgroundColorAttr    = "106"
+	brightWhiteBackgroundColorAttr   = "107"
+)
+
+// foregroundColorString returns the style SGR attribute for the given
+// foreground color.
+// See: https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_(Select_Graphic_Rendition)_parameters
+func foregroundColorString(c Color) string {
+	switch c := c.(type) {
+	case BasicColor:
+		// 3-bit or 4-bit ANSI foreground
+		// "3<n>" or "9<n>" where n is the color number from 0 to 7
+		switch c {
+		case Black:
+			return blackForegroundColorAttr
+		case Red:
+			return redForegroundColorAttr
+		case Green:
+			return greenForegroundColorAttr
+		case Yellow:
+			return yellowForegroundColorAttr
+		case Blue:
+			return blueForegroundColorAttr
+		case Magenta:
+			return magentaForegroundColorAttr
+		case Cyan:
+			return cyanForegroundColorAttr
+		case White:
+			return whiteForegroundColorAttr
+		case BrightBlack:
+			return brightBlackForegroundColorAttr
+		case BrightRed:
+			return brightRedForegroundColorAttr
+		case BrightGreen:
+			return brightGreenForegroundColorAttr
+		case BrightYellow:
+			return brightYellowForegroundColorAttr
+		case BrightBlue:
+			return brightBlueForegroundColorAttr
+		case BrightMagenta:
+			return brightMagentaForegroundColorAttr
+		case BrightCyan:
+			return brightCyanForegroundColorAttr
+		case BrightWhite:
+			return brightWhiteForegroundColorAttr
+		}
+	case ExtendedColor:
+		// 256-color ANSI foreground
+		// "38;5;<n>"
+		return "38;5;" + strconv.FormatUint(uint64(c), 10)
+	case TrueColor, color.Color:
+		// 24-bit "true color" foreground
+		// "38;2;<r>;<g>;<b>"
+		r, g, b, _ := c.RGBA()
+		return "38;2;" +
+			strconv.FormatUint(uint64(shift(r)), 10) + ";" +
+			strconv.FormatUint(uint64(shift(g)), 10) + ";" +
+			strconv.FormatUint(uint64(shift(b)), 10)
+	}
+	return defaultForegroundColorAttr
+}
+
+// backgroundColorString returns the style SGR attribute for the given
+// background color.
+// See: https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_(Select_Graphic_Rendition)_parameters
+func backgroundColorString(c Color) string {
+	switch c := c.(type) {
+	case BasicColor:
+		// 3-bit or 4-bit ANSI foreground
+		// "4<n>" or "10<n>" where n is the color number from 0 to 7
+		switch c {
+		case Black:
+			return blackBackgroundColorAttr
+		case Red:
+			return redBackgroundColorAttr
+		case Green:
+			return greenBackgroundColorAttr
+		case Yellow:
+			return yellowBackgroundColorAttr
+		case Blue:
+			return blueBackgroundColorAttr
+		case Magenta:
+			return magentaBackgroundColorAttr
+		case Cyan:
+			return cyanBackgroundColorAttr
+		case White:
+			return whiteBackgroundColorAttr
+		case BrightBlack:
+			return brightBlackBackgroundColorAttr
+		case BrightRed:
+			return brightRedBackgroundColorAttr
+		case BrightGreen:
+			return brightGreenBackgroundColorAttr
+		case BrightYellow:
+			return brightYellowBackgroundColorAttr
+		case BrightBlue:
+			return brightBlueBackgroundColorAttr
+		case BrightMagenta:
+			return brightMagentaBackgroundColorAttr
+		case BrightCyan:
+			return brightCyanBackgroundColorAttr
+		case BrightWhite:
+			return brightWhiteBackgroundColorAttr
+		}
+	case ExtendedColor:
+		// 256-color ANSI foreground
+		// "48;5;<n>"
+		return "48;5;" + strconv.FormatUint(uint64(c), 10)
+	case TrueColor, color.Color:
+		// 24-bit "true color" foreground
+		// "38;2;<r>;<g>;<b>"
+		r, g, b, _ := c.RGBA()
+		return "48;2;" +
+			strconv.FormatUint(uint64(shift(r)), 10) + ";" +
+			strconv.FormatUint(uint64(shift(g)), 10) + ";" +
+			strconv.FormatUint(uint64(shift(b)), 10)
+	}
+	return defaultBackgroundColorAttr
+}
+
+// underlineColorString returns the style SGR attribute for the given underline
+// color.
+// See: https://en.wikipedia.org/wiki/ANSI_escape_code#SGR_(Select_Graphic_Rendition)_parameters
+func underlineColorString(c Color) string {
+	switch c := c.(type) {
+	// NOTE: we can't use 3-bit and 4-bit ANSI color codes with underline
+	// color, use 256-color instead.
+	//
+	// 256-color ANSI underline color
+	// "58;5;<n>"
+	case BasicColor:
+		return "58;5;" + strconv.FormatUint(uint64(c), 10)
+	case ExtendedColor:
+		return "58;5;" + strconv.FormatUint(uint64(c), 10)
+	case TrueColor, color.Color:
+		// 24-bit "true color" foreground
+		// "38;2;<r>;<g>;<b>"
+		r, g, b, _ := c.RGBA()
+		return "58;2;" +
+			strconv.FormatUint(uint64(shift(r)), 10) + ";" +
+			strconv.FormatUint(uint64(shift(g)), 10) + ";" +
+			strconv.FormatUint(uint64(shift(b)), 10)
+	}
+	return defaultUnderlineColorAttr
+}
+
+// ReadStyleColor decodes a color from a slice of parameters. It returns the
+// number of parameters read and the color. This function is used to read SGR
+// color parameters following the ITU T.416 standard.
+//
+// It supports reading the following color types:
+//   - 0: implementation defined
+//   - 1: transparent
+//   - 2: RGB direct color
+//   - 3: CMY direct color
+//   - 4: CMYK direct color
+//   - 5: indexed color
+//   - 6: RGBA direct color (WezTerm extension)
+//
+// The parameters can be separated by semicolons (;) or colons (:). Mixing
+// separators is not allowed.
+//
+// The specs supports defining a color space id, a color tolerance value, and a
+// tolerance color space id. However, these values have no effect on the
+// returned color and will be ignored.
+//
+// This implementation includes a few modifications to the specs:
+//  1. Support for legacy color values separated by semicolons (;) with respect to RGB, and indexed colors
+//  2. Support ignoring and omitting the color space id (second parameter) with respect to RGB colors
+//  3. Support ignoring and omitting the 6th parameter with respect to RGB and CMY colors
+//  4. Support reading RGBA colors
+func ReadStyleColor(params Params, co *color.Color) (n int) {
+	if len(params) < 2 { // Need at least SGR type and color type
+		return 0
+	}
+
+	// First parameter indicates one of 38, 48, or 58 (foreground, background, or underline)
+	s := params[0]
+	p := params[1]
+	colorType := p.Param(0)
+	n = 2
+
+	paramsfn := func() (p1, p2, p3, p4 int) {
+		// Where should we start reading the color?
+		switch {
+		case s.HasMore() && p.HasMore() && len(params) > 8 && params[2].HasMore() && params[3].HasMore() && params[4].HasMore() && params[5].HasMore() && params[6].HasMore() && params[7].HasMore():
+			// We have color space id, a 6th parameter, a tolerance value, and a tolerance color space
+			n += 7
+			return params[3].Param(0), params[4].Param(0), params[5].Param(0), params[6].Param(0)
+		case s.HasMore() && p.HasMore() && len(params) > 7 && params[2].HasMore() && params[3].HasMore() && params[4].HasMore() && params[5].HasMore() && params[6].HasMore():
+			// We have color space id, a 6th parameter, and a tolerance value
+			n += 6
+			return params[3].Param(0), params[4].Param(0), params[5].Param(0), params[6].Param(0)
+		case s.HasMore() && p.HasMore() && len(params) > 6 && params[2].HasMore() && params[3].HasMore() && params[4].HasMore() && params[5].HasMore():
+			// We have color space id and a 6th parameter
+			// 48 : 4 : : 1 : 2 : 3 :4
+			n += 5
+			return params[3].Param(0), params[4].Param(0), params[5].Param(0), params[6].Param(0)
+		case s.HasMore() && p.HasMore() && len(params) > 5 && params[2].HasMore() && params[3].HasMore() && params[4].HasMore() && !params[5].HasMore():
+			// We have color space
+			// 48 : 3 : : 1 : 2 : 3
+			n += 4
+			return params[3].Param(0), params[4].Param(0), params[5].Param(0), -1
+		case s.HasMore() && p.HasMore() && p.Param(0) == 2 && params[2].HasMore() && params[3].HasMore() && !params[4].HasMore():
+			// We have color values separated by colons (:)
+			// 48 : 2 : 1 : 2 : 3
+			fallthrough
+		case !s.HasMore() && !p.HasMore() && p.Param(0) == 2 && !params[2].HasMore() && !params[3].HasMore() && !params[4].HasMore():
+			// Support legacy color values separated by semicolons (;)
+			// 48 ; 2 ; 1 ; 2 ; 3
+			n += 3
+			return params[2].Param(0), params[3].Param(0), params[4].Param(0), -1
+		}
+		// Ambiguous SGR color
+		return -1, -1, -1, -1
+	}
+
+	switch colorType {
+	case 0: // implementation defined
+		return 2
+	case 1: // transparent
+		*co = color.Transparent
+		return 2
+	case 2: // RGB direct color
+		if len(params) < 5 {
+			return 0
+		}
+
+		r, g, b, _ := paramsfn()
+		if r == -1 || g == -1 || b == -1 {
+			return 0
+		}
+
+		*co = color.RGBA{
+			R: uint8(r), //nolint:gosec
+			G: uint8(g), //nolint:gosec
+			B: uint8(b), //nolint:gosec
+			A: 0xff,
+		}
+		return
+
+	case 3: // CMY direct color
+		if len(params) < 5 {
+			return 0
+		}
+
+		c, m, y, _ := paramsfn()
+		if c == -1 || m == -1 || y == -1 {
+			return 0
+		}
+
+		*co = color.CMYK{
+			C: uint8(c), //nolint:gosec
+			M: uint8(m), //nolint:gosec
+			Y: uint8(y), //nolint:gosec
+			K: 0,
+		}
+		return
+
+	case 4: // CMYK direct color
+		if len(params) < 6 {
+			return 0
+		}
+
+		c, m, y, k := paramsfn()
+		if c == -1 || m == -1 || y == -1 || k == -1 {
+			return 0
+		}
+
+		*co = color.CMYK{
+			C: uint8(c), //nolint:gosec
+			M: uint8(m), //nolint:gosec
+			Y: uint8(y), //nolint:gosec
+			K: uint8(k), //nolint:gosec
+		}
+		return
+
+	case 5: // indexed color
+		if len(params) < 3 {
+			return 0
+		}
+		switch {
+		case s.HasMore() && p.HasMore() && !params[2].HasMore():
+			// Colon separated indexed color
+			// 38 : 5 : 234
+		case !s.HasMore() && !p.HasMore() && !params[2].HasMore():
+			// Legacy semicolon indexed color
+			// 38 ; 5 ; 234
+		default:
+			return 0
+		}
+		*co = ExtendedColor(params[2].Param(0)) //nolint:gosec
+		return 3
+
+	case 6: // RGBA direct color
+		if len(params) < 6 {
+			return 0
+		}
+
+		r, g, b, a := paramsfn()
+		if r == -1 || g == -1 || b == -1 || a == -1 {
+			return 0
+		}
+
+		*co = color.RGBA{
+			R: uint8(r), //nolint:gosec
+			G: uint8(g), //nolint:gosec
+			B: uint8(b), //nolint:gosec
+			A: uint8(a), //nolint:gosec
+		}
+		return
+
+	default:
+		return 0
+	}
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/termcap.go b/vendor/github.com/charmbracelet/x/ansi/termcap.go
new file mode 100644
index 0000000..3c5c7da
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/termcap.go
@@ -0,0 +1,41 @@
+package ansi
+
+import (
+	"encoding/hex"
+	"strings"
+)
+
+// RequestTermcap (XTGETTCAP) requests Termcap/Terminfo strings.
+//
+//	DCS + q <Pt> ST
+//
+// Where <Pt> is a list of Termcap/Terminfo capabilities, encoded in 2-digit
+// hexadecimals, separated by semicolons.
+//
+// See: https://man7.org/linux/man-pages/man5/terminfo.5.html
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Operating-System-Commands
+func XTGETTCAP(caps ...string) string {
+	if len(caps) == 0 {
+		return ""
+	}
+
+	s := "\x1bP+q"
+	for i, c := range caps {
+		if i > 0 {
+			s += ";"
+		}
+		s += strings.ToUpper(hex.EncodeToString([]byte(c)))
+	}
+
+	return s + "\x1b\\"
+}
+
+// RequestTermcap is an alias for [XTGETTCAP].
+func RequestTermcap(caps ...string) string {
+	return XTGETTCAP(caps...)
+}
+
+// RequestTerminfo is an alias for [XTGETTCAP].
+func RequestTerminfo(caps ...string) string {
+	return XTGETTCAP(caps...)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/title.go b/vendor/github.com/charmbracelet/x/ansi/title.go
new file mode 100644
index 0000000..8fd8bf9
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/title.go
@@ -0,0 +1,32 @@
+package ansi
+
+// SetIconNameWindowTitle returns a sequence for setting the icon name and
+// window title.
+//
+//	OSC 0 ; title ST
+//	OSC 0 ; title BEL
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Operating-System-Commands
+func SetIconNameWindowTitle(s string) string {
+	return "\x1b]0;" + s + "\x07"
+}
+
+// SetIconName returns a sequence for setting the icon name.
+//
+//	OSC 1 ; title ST
+//	OSC 1 ; title BEL
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Operating-System-Commands
+func SetIconName(s string) string {
+	return "\x1b]1;" + s + "\x07"
+}
+
+// SetWindowTitle returns a sequence for setting the window title.
+//
+//	OSC 2 ; title ST
+//	OSC 2 ; title BEL
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h2-Operating-System-Commands
+func SetWindowTitle(s string) string {
+	return "\x1b]2;" + s + "\x07"
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/truncate.go b/vendor/github.com/charmbracelet/x/ansi/truncate.go
new file mode 100644
index 0000000..1fa3efe
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/truncate.go
@@ -0,0 +1,282 @@
+package ansi
+
+import (
+	"bytes"
+
+	"github.com/charmbracelet/x/ansi/parser"
+	"github.com/mattn/go-runewidth"
+	"github.com/rivo/uniseg"
+)
+
+// Cut the string, without adding any prefix or tail strings. This function is
+// aware of ANSI escape codes and will not break them, and accounts for
+// wide-characters (such as East-Asian characters and emojis). Note that the
+// [left] parameter is inclusive, while [right] isn't.
+// This treats the text as a sequence of graphemes.
+func Cut(s string, left, right int) string {
+	return cut(GraphemeWidth, s, left, right)
+}
+
+// CutWc the string, without adding any prefix or tail strings. This function is
+// aware of ANSI escape codes and will not break them, and accounts for
+// wide-characters (such as East-Asian characters and emojis). Note that the
+// [left] parameter is inclusive, while [right] isn't.
+// This treats the text as a sequence of wide characters and runes.
+func CutWc(s string, left, right int) string {
+	return cut(WcWidth, s, left, right)
+}
+
+func cut(m Method, s string, left, right int) string {
+	if right <= left {
+		return ""
+	}
+
+	truncate := Truncate
+	truncateLeft := TruncateLeft
+	if m == WcWidth {
+		truncate = TruncateWc
+		truncateLeft = TruncateWc
+	}
+
+	if left == 0 {
+		return truncate(s, right, "")
+	}
+	return truncateLeft(Truncate(s, right, ""), left, "")
+}
+
+// Truncate truncates a string to a given length, adding a tail to the end if
+// the string is longer than the given length. This function is aware of ANSI
+// escape codes and will not break them, and accounts for wide-characters (such
+// as East-Asian characters and emojis).
+// This treats the text as a sequence of graphemes.
+func Truncate(s string, length int, tail string) string {
+	return truncate(GraphemeWidth, s, length, tail)
+}
+
+// TruncateWc truncates a string to a given length, adding a tail to the end if
+// the string is longer than the given length. This function is aware of ANSI
+// escape codes and will not break them, and accounts for wide-characters (such
+// as East-Asian characters and emojis).
+// This treats the text as a sequence of wide characters and runes.
+func TruncateWc(s string, length int, tail string) string {
+	return truncate(WcWidth, s, length, tail)
+}
+
+func truncate(m Method, s string, length int, tail string) string {
+	if sw := StringWidth(s); sw <= length {
+		return s
+	}
+
+	tw := StringWidth(tail)
+	length -= tw
+	if length < 0 {
+		return ""
+	}
+
+	var cluster []byte
+	var buf bytes.Buffer
+	curWidth := 0
+	ignoring := false
+	pstate := parser.GroundState // initial state
+	b := []byte(s)
+	i := 0
+
+	// Here we iterate over the bytes of the string and collect printable
+	// characters and runes. We also keep track of the width of the string
+	// in cells.
+	//
+	// Once we reach the given length, we start ignoring characters and only
+	// collect ANSI escape codes until we reach the end of string.
+	for i < len(b) {
+		state, action := parser.Table.Transition(pstate, b[i])
+		if state == parser.Utf8State {
+			// This action happens when we transition to the Utf8State.
+			var width int
+			cluster, _, width, _ = uniseg.FirstGraphemeCluster(b[i:], -1)
+			if m == WcWidth {
+				width = runewidth.StringWidth(string(cluster))
+			}
+
+			// increment the index by the length of the cluster
+			i += len(cluster)
+
+			// Are we ignoring? Skip to the next byte
+			if ignoring {
+				continue
+			}
+
+			// Is this gonna be too wide?
+			// If so write the tail and stop collecting.
+			if curWidth+width > length && !ignoring {
+				ignoring = true
+				buf.WriteString(tail)
+			}
+
+			if curWidth+width > length {
+				continue
+			}
+
+			curWidth += width
+			buf.Write(cluster)
+
+			// Done collecting, now we're back in the ground state.
+			pstate = parser.GroundState
+			continue
+		}
+
+		switch action {
+		case parser.PrintAction:
+			// Is this gonna be too wide?
+			// If so write the tail and stop collecting.
+			if curWidth >= length && !ignoring {
+				ignoring = true
+				buf.WriteString(tail)
+			}
+
+			// Skip to the next byte if we're ignoring
+			if ignoring {
+				i++
+				continue
+			}
+
+			// collects printable ASCII
+			curWidth++
+			fallthrough
+		default:
+			buf.WriteByte(b[i])
+			i++
+		}
+
+		// Transition to the next state.
+		pstate = state
+
+		// Once we reach the given length, we start ignoring runes and write
+		// the tail to the buffer.
+		if curWidth > length && !ignoring {
+			ignoring = true
+			buf.WriteString(tail)
+		}
+	}
+
+	return buf.String()
+}
+
+// TruncateLeft truncates a string from the left side by removing n characters,
+// adding a prefix to the beginning if the string is longer than n.
+// This function is aware of ANSI escape codes and will not break them, and
+// accounts for wide-characters (such as East-Asian characters and emojis).
+// This treats the text as a sequence of graphemes.
+func TruncateLeft(s string, n int, prefix string) string {
+	return truncateLeft(GraphemeWidth, s, n, prefix)
+}
+
+// TruncateLeftWc truncates a string from the left side by removing n characters,
+// adding a prefix to the beginning if the string is longer than n.
+// This function is aware of ANSI escape codes and will not break them, and
+// accounts for wide-characters (such as East-Asian characters and emojis).
+// This treats the text as a sequence of wide characters and runes.
+func TruncateLeftWc(s string, n int, prefix string) string {
+	return truncateLeft(WcWidth, s, n, prefix)
+}
+
+func truncateLeft(m Method, s string, n int, prefix string) string {
+	if n <= 0 {
+		return s
+	}
+
+	var cluster []byte
+	var buf bytes.Buffer
+	curWidth := 0
+	ignoring := true
+	pstate := parser.GroundState
+	b := []byte(s)
+	i := 0
+
+	for i < len(b) {
+		if !ignoring {
+			buf.Write(b[i:])
+			break
+		}
+
+		state, action := parser.Table.Transition(pstate, b[i])
+		if state == parser.Utf8State {
+			var width int
+			cluster, _, width, _ = uniseg.FirstGraphemeCluster(b[i:], -1)
+			if m == WcWidth {
+				width = runewidth.StringWidth(string(cluster))
+			}
+
+			i += len(cluster)
+			curWidth += width
+
+			if curWidth > n && ignoring {
+				ignoring = false
+				buf.WriteString(prefix)
+			}
+
+			if ignoring {
+				continue
+			}
+
+			if curWidth > n {
+				buf.Write(cluster)
+			}
+
+			pstate = parser.GroundState
+			continue
+		}
+
+		switch action {
+		case parser.PrintAction:
+			curWidth++
+
+			if curWidth > n && ignoring {
+				ignoring = false
+				buf.WriteString(prefix)
+			}
+
+			if ignoring {
+				i++
+				continue
+			}
+
+			fallthrough
+		default:
+			buf.WriteByte(b[i])
+			i++
+		}
+
+		pstate = state
+		if curWidth > n && ignoring {
+			ignoring = false
+			buf.WriteString(prefix)
+		}
+	}
+
+	return buf.String()
+}
+
+// ByteToGraphemeRange takes start and stop byte positions and converts them to
+// grapheme-aware char positions.
+// You can use this with [Truncate], [TruncateLeft], and [Cut].
+func ByteToGraphemeRange(str string, byteStart, byteStop int) (charStart, charStop int) {
+	bytePos, charPos := 0, 0
+	gr := uniseg.NewGraphemes(str)
+	for byteStart > bytePos {
+		if !gr.Next() {
+			break
+		}
+		bytePos += len(gr.Str())
+		charPos += max(1, gr.Width())
+	}
+	charStart = charPos
+	for byteStop > bytePos {
+		if !gr.Next() {
+			break
+		}
+		bytePos += len(gr.Str())
+		charPos += max(1, gr.Width())
+	}
+	charStop = charPos
+	return
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/util.go b/vendor/github.com/charmbracelet/x/ansi/util.go
new file mode 100644
index 0000000..301ef15
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/util.go
@@ -0,0 +1,106 @@
+package ansi
+
+import (
+	"fmt"
+	"image/color"
+	"strconv"
+	"strings"
+
+	"github.com/lucasb-eyer/go-colorful"
+)
+
+// colorToHexString returns a hex string representation of a color.
+func colorToHexString(c color.Color) string {
+	if c == nil {
+		return ""
+	}
+	shift := func(v uint32) uint32 {
+		if v > 0xff {
+			return v >> 8
+		}
+		return v
+	}
+	r, g, b, _ := c.RGBA()
+	r, g, b = shift(r), shift(g), shift(b)
+	return fmt.Sprintf("#%02x%02x%02x", r, g, b)
+}
+
+// rgbToHex converts red, green, and blue values to a hexadecimal value.
+//
+//	hex := rgbToHex(0, 0, 255) // 0x0000FF
+func rgbToHex(r, g, b uint32) uint32 {
+	return r<<16 + g<<8 + b
+}
+
+type shiftable interface {
+	~uint | ~uint16 | ~uint32 | ~uint64
+}
+
+func shift[T shiftable](x T) T {
+	if x > 0xff {
+		x >>= 8
+	}
+	return x
+}
+
+// XParseColor is a helper function that parses a string into a color.Color. It
+// provides a similar interface to the XParseColor function in Xlib. It
+// supports the following formats:
+//
+//   - #RGB
+//   - #RRGGBB
+//   - rgb:RRRR/GGGG/BBBB
+//   - rgba:RRRR/GGGG/BBBB/AAAA
+//
+// If the string is not a valid color, nil is returned.
+//
+// See: https://linux.die.net/man/3/xparsecolor
+func XParseColor(s string) color.Color {
+	switch {
+	case strings.HasPrefix(s, "#"):
+		c, err := colorful.Hex(s)
+		if err != nil {
+			return nil
+		}
+
+		return c
+	case strings.HasPrefix(s, "rgb:"):
+		parts := strings.Split(s[4:], "/")
+		if len(parts) != 3 {
+			return nil
+		}
+
+		r, _ := strconv.ParseUint(parts[0], 16, 32)
+		g, _ := strconv.ParseUint(parts[1], 16, 32)
+		b, _ := strconv.ParseUint(parts[2], 16, 32)
+
+		return color.RGBA{uint8(shift(r)), uint8(shift(g)), uint8(shift(b)), 255} //nolint:gosec
+	case strings.HasPrefix(s, "rgba:"):
+		parts := strings.Split(s[5:], "/")
+		if len(parts) != 4 {
+			return nil
+		}
+
+		r, _ := strconv.ParseUint(parts[0], 16, 32)
+		g, _ := strconv.ParseUint(parts[1], 16, 32)
+		b, _ := strconv.ParseUint(parts[2], 16, 32)
+		a, _ := strconv.ParseUint(parts[3], 16, 32)
+
+		return color.RGBA{uint8(shift(r)), uint8(shift(g)), uint8(shift(b)), uint8(shift(a))} //nolint:gosec
+	}
+	return nil
+}
+
+type ordered interface {
+	~int | ~int8 | ~int16 | ~int32 | ~int64 |
+		~uint | ~uint8 | ~uint16 | ~uint32 | ~uint64 | ~uintptr |
+		~float32 | ~float64 |
+		~string
+}
+
+func max[T ordered](a, b T) T { //nolint:predeclared
+	if a > b {
+		return a
+	}
+	return b
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/width.go b/vendor/github.com/charmbracelet/x/ansi/width.go
new file mode 100644
index 0000000..d0487d3
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/width.go
@@ -0,0 +1,113 @@
+package ansi
+
+import (
+	"bytes"
+
+	"github.com/charmbracelet/x/ansi/parser"
+	"github.com/mattn/go-runewidth"
+	"github.com/rivo/uniseg"
+)
+
+// Strip removes ANSI escape codes from a string.
+func Strip(s string) string {
+	var (
+		buf    bytes.Buffer         // buffer for collecting printable characters
+		ri     int                  // rune index
+		rw     int                  // rune width
+		pstate = parser.GroundState // initial state
+	)
+
+	// This implements a subset of the Parser to only collect runes and
+	// printable characters.
+	for i := 0; i < len(s); i++ {
+		if pstate == parser.Utf8State {
+			// During this state, collect rw bytes to form a valid rune in the
+			// buffer. After getting all the rune bytes into the buffer,
+			// transition to GroundState and reset the counters.
+			buf.WriteByte(s[i])
+			ri++
+			if ri < rw {
+				continue
+			}
+			pstate = parser.GroundState
+			ri = 0
+			rw = 0
+			continue
+		}
+
+		state, action := parser.Table.Transition(pstate, s[i])
+		switch action {
+		case parser.CollectAction:
+			if state == parser.Utf8State {
+				// This action happens when we transition to the Utf8State.
+				rw = utf8ByteLen(s[i])
+				buf.WriteByte(s[i])
+				ri++
+			}
+		case parser.PrintAction, parser.ExecuteAction:
+			// collects printable ASCII and non-printable characters
+			buf.WriteByte(s[i])
+		}
+
+		// Transition to the next state.
+		// The Utf8State is managed separately above.
+		if pstate != parser.Utf8State {
+			pstate = state
+		}
+	}
+
+	return buf.String()
+}
+
+// StringWidth returns the width of a string in cells. This is the number of
+// cells that the string will occupy when printed in a terminal. ANSI escape
+// codes are ignored and wide characters (such as East Asians and emojis) are
+// accounted for.
+// This treats the text as a sequence of grapheme clusters.
+func StringWidth(s string) int {
+	return stringWidth(GraphemeWidth, s)
+}
+
+// StringWidthWc returns the width of a string in cells. This is the number of
+// cells that the string will occupy when printed in a terminal. ANSI escape
+// codes are ignored and wide characters (such as East Asians and emojis) are
+// accounted for.
+// This treats the text as a sequence of wide characters and runes.
+func StringWidthWc(s string) int {
+	return stringWidth(WcWidth, s)
+}
+
+func stringWidth(m Method, s string) int {
+	if s == "" {
+		return 0
+	}
+
+	var (
+		pstate  = parser.GroundState // initial state
+		cluster string
+		width   int
+	)
+
+	for i := 0; i < len(s); i++ {
+		state, action := parser.Table.Transition(pstate, s[i])
+		if state == parser.Utf8State {
+			var w int
+			cluster, _, w, _ = uniseg.FirstGraphemeClusterInString(s[i:], -1)
+			if m == WcWidth {
+				w = runewidth.StringWidth(cluster)
+			}
+			width += w
+			i += len(cluster) - 1
+			pstate = parser.GroundState
+			continue
+		}
+
+		if action == parser.PrintAction {
+			width++
+		}
+
+		pstate = state
+	}
+
+	return width
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/winop.go b/vendor/github.com/charmbracelet/x/ansi/winop.go
new file mode 100644
index 0000000..0238780
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/winop.go
@@ -0,0 +1,53 @@
+package ansi
+
+import (
+	"strconv"
+	"strings"
+)
+
+const (
+	// ResizeWindowWinOp is a window operation that resizes the terminal
+	// window.
+	ResizeWindowWinOp = 4
+
+	// RequestWindowSizeWinOp is a window operation that requests a report of
+	// the size of the terminal window in pixels. The response is in the form:
+	//  CSI 4 ; height ; width t
+	RequestWindowSizeWinOp = 14
+
+	// RequestCellSizeWinOp is a window operation that requests a report of
+	// the size of the terminal cell size in pixels. The response is in the form:
+	//  CSI 6 ; height ; width t
+	RequestCellSizeWinOp = 16
+)
+
+// WindowOp (XTWINOPS) is a sequence that manipulates the terminal window.
+//
+//	CSI Ps ; Ps ; Ps t
+//
+// Ps is a semicolon-separated list of parameters.
+// See https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h4-Functions-using-CSI-_-ordered-by-the-final-character-lparen-s-rparen:CSI-Ps;Ps;Ps-t.1EB0
+func WindowOp(p int, ps ...int) string {
+	if p <= 0 {
+		return ""
+	}
+
+	if len(ps) == 0 {
+		return "\x1b[" + strconv.Itoa(p) + "t"
+	}
+
+	params := make([]string, 0, len(ps)+1)
+	params = append(params, strconv.Itoa(p))
+	for _, p := range ps {
+		if p >= 0 {
+			params = append(params, strconv.Itoa(p))
+		}
+	}
+
+	return "\x1b[" + strings.Join(params, ";") + "t"
+}
+
+// XTWINOPS is an alias for [WindowOp].
+func XTWINOPS(p int, ps ...int) string {
+	return WindowOp(p, ps...)
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/wrap.go b/vendor/github.com/charmbracelet/x/ansi/wrap.go
new file mode 100644
index 0000000..6b99580
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/wrap.go
@@ -0,0 +1,467 @@
+package ansi
+
+import (
+	"bytes"
+	"unicode"
+	"unicode/utf8"
+
+	"github.com/charmbracelet/x/ansi/parser"
+	"github.com/mattn/go-runewidth"
+	"github.com/rivo/uniseg"
+)
+
+// nbsp is a non-breaking space
+const nbsp = 0xA0
+
+// Hardwrap wraps a string or a block of text to a given line length, breaking
+// word boundaries. This will preserve ANSI escape codes and will account for
+// wide-characters in the string.
+// When preserveSpace is true, spaces at the beginning of a line will be
+// preserved.
+// This treats the text as a sequence of graphemes.
+func Hardwrap(s string, limit int, preserveSpace bool) string {
+	return hardwrap(GraphemeWidth, s, limit, preserveSpace)
+}
+
+// HardwrapWc wraps a string or a block of text to a given line length, breaking
+// word boundaries. This will preserve ANSI escape codes and will account for
+// wide-characters in the string.
+// When preserveSpace is true, spaces at the beginning of a line will be
+// preserved.
+// This treats the text as a sequence of wide characters and runes.
+func HardwrapWc(s string, limit int, preserveSpace bool) string {
+	return hardwrap(WcWidth, s, limit, preserveSpace)
+}
+
+func hardwrap(m Method, s string, limit int, preserveSpace bool) string {
+	if limit < 1 {
+		return s
+	}
+
+	var (
+		cluster      []byte
+		buf          bytes.Buffer
+		curWidth     int
+		forceNewline bool
+		pstate       = parser.GroundState // initial state
+		b            = []byte(s)
+	)
+
+	addNewline := func() {
+		buf.WriteByte('\n')
+		curWidth = 0
+	}
+
+	i := 0
+	for i < len(b) {
+		state, action := parser.Table.Transition(pstate, b[i])
+		if state == parser.Utf8State {
+			var width int
+			cluster, _, width, _ = uniseg.FirstGraphemeCluster(b[i:], -1)
+			if m == WcWidth {
+				width = runewidth.StringWidth(string(cluster))
+			}
+			i += len(cluster)
+
+			if curWidth+width > limit {
+				addNewline()
+			}
+			if !preserveSpace && curWidth == 0 && len(cluster) <= 4 {
+				// Skip spaces at the beginning of a line
+				if r, _ := utf8.DecodeRune(cluster); r != utf8.RuneError && unicode.IsSpace(r) {
+					pstate = parser.GroundState
+					continue
+				}
+			}
+
+			buf.Write(cluster)
+			curWidth += width
+			pstate = parser.GroundState
+			continue
+		}
+
+		switch action {
+		case parser.PrintAction, parser.ExecuteAction:
+			if b[i] == '\n' {
+				addNewline()
+				forceNewline = false
+				break
+			}
+
+			if curWidth+1 > limit {
+				addNewline()
+				forceNewline = true
+			}
+
+			// Skip spaces at the beginning of a line
+			if curWidth == 0 {
+				if !preserveSpace && forceNewline && unicode.IsSpace(rune(b[i])) {
+					break
+				}
+				forceNewline = false
+			}
+
+			buf.WriteByte(b[i])
+			if action == parser.PrintAction {
+				curWidth++
+			}
+		default:
+			buf.WriteByte(b[i])
+		}
+
+		// We manage the UTF8 state separately manually above.
+		if pstate != parser.Utf8State {
+			pstate = state
+		}
+		i++
+	}
+
+	return buf.String()
+}
+
+// Wordwrap wraps a string or a block of text to a given line length, not
+// breaking word boundaries. This will preserve ANSI escape codes and will
+// account for wide-characters in the string.
+// The breakpoints string is a list of characters that are considered
+// breakpoints for word wrapping. A hyphen (-) is always considered a
+// breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+//
+// This treats the text as a sequence of graphemes.
+func Wordwrap(s string, limit int, breakpoints string) string {
+	return wordwrap(GraphemeWidth, s, limit, breakpoints)
+}
+
+// WordwrapWc wraps a string or a block of text to a given line length, not
+// breaking word boundaries. This will preserve ANSI escape codes and will
+// account for wide-characters in the string.
+// The breakpoints string is a list of characters that are considered
+// breakpoints for word wrapping. A hyphen (-) is always considered a
+// breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+//
+// This treats the text as a sequence of wide characters and runes.
+func WordwrapWc(s string, limit int, breakpoints string) string {
+	return wordwrap(WcWidth, s, limit, breakpoints)
+}
+
+func wordwrap(m Method, s string, limit int, breakpoints string) string {
+	if limit < 1 {
+		return s
+	}
+
+	var (
+		cluster  []byte
+		buf      bytes.Buffer
+		word     bytes.Buffer
+		space    bytes.Buffer
+		curWidth int
+		wordLen  int
+		pstate   = parser.GroundState // initial state
+		b        = []byte(s)
+	)
+
+	addSpace := func() {
+		curWidth += space.Len()
+		buf.Write(space.Bytes())
+		space.Reset()
+	}
+
+	addWord := func() {
+		if word.Len() == 0 {
+			return
+		}
+
+		addSpace()
+		curWidth += wordLen
+		buf.Write(word.Bytes())
+		word.Reset()
+		wordLen = 0
+	}
+
+	addNewline := func() {
+		buf.WriteByte('\n')
+		curWidth = 0
+		space.Reset()
+	}
+
+	i := 0
+	for i < len(b) {
+		state, action := parser.Table.Transition(pstate, b[i])
+		if state == parser.Utf8State {
+			var width int
+			cluster, _, width, _ = uniseg.FirstGraphemeCluster(b[i:], -1)
+			if m == WcWidth {
+				width = runewidth.StringWidth(string(cluster))
+			}
+			i += len(cluster)
+
+			r, _ := utf8.DecodeRune(cluster)
+			if r != utf8.RuneError && unicode.IsSpace(r) && r != nbsp {
+				addWord()
+				space.WriteRune(r)
+			} else if bytes.ContainsAny(cluster, breakpoints) {
+				addSpace()
+				addWord()
+				buf.Write(cluster)
+				curWidth++
+			} else {
+				word.Write(cluster)
+				wordLen += width
+				if curWidth+space.Len()+wordLen > limit &&
+					wordLen < limit {
+					addNewline()
+				}
+			}
+
+			pstate = parser.GroundState
+			continue
+		}
+
+		switch action {
+		case parser.PrintAction, parser.ExecuteAction:
+			r := rune(b[i])
+			switch {
+			case r == '\n':
+				if wordLen == 0 {
+					if curWidth+space.Len() > limit {
+						curWidth = 0
+					} else {
+						buf.Write(space.Bytes())
+					}
+					space.Reset()
+				}
+
+				addWord()
+				addNewline()
+			case unicode.IsSpace(r):
+				addWord()
+				space.WriteByte(b[i])
+			case r == '-':
+				fallthrough
+			case runeContainsAny(r, breakpoints):
+				addSpace()
+				addWord()
+				buf.WriteByte(b[i])
+				curWidth++
+			default:
+				word.WriteByte(b[i])
+				wordLen++
+				if curWidth+space.Len()+wordLen > limit &&
+					wordLen < limit {
+					addNewline()
+				}
+			}
+
+		default:
+			word.WriteByte(b[i])
+		}
+
+		// We manage the UTF8 state separately manually above.
+		if pstate != parser.Utf8State {
+			pstate = state
+		}
+		i++
+	}
+
+	addWord()
+
+	return buf.String()
+}
+
+// Wrap wraps a string or a block of text to a given line length, breaking word
+// boundaries if necessary. This will preserve ANSI escape codes and will
+// account for wide-characters in the string. The breakpoints string is a list
+// of characters that are considered breakpoints for word wrapping. A hyphen
+// (-) is always considered a breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+//
+// This treats the text as a sequence of graphemes.
+func Wrap(s string, limit int, breakpoints string) string {
+	return wrap(GraphemeWidth, s, limit, breakpoints)
+}
+
+// WrapWc wraps a string or a block of text to a given line length, breaking word
+// boundaries if necessary. This will preserve ANSI escape codes and will
+// account for wide-characters in the string. The breakpoints string is a list
+// of characters that are considered breakpoints for word wrapping. A hyphen
+// (-) is always considered a breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+//
+// This treats the text as a sequence of wide characters and runes.
+func WrapWc(s string, limit int, breakpoints string) string {
+	return wrap(WcWidth, s, limit, breakpoints)
+}
+
+func wrap(m Method, s string, limit int, breakpoints string) string {
+	if limit < 1 {
+		return s
+	}
+
+	var (
+		cluster  []byte
+		buf      bytes.Buffer
+		word     bytes.Buffer
+		space    bytes.Buffer
+		curWidth int                  // written width of the line
+		wordLen  int                  // word buffer len without ANSI escape codes
+		pstate   = parser.GroundState // initial state
+		b        = []byte(s)
+	)
+
+	addSpace := func() {
+		curWidth += space.Len()
+		buf.Write(space.Bytes())
+		space.Reset()
+	}
+
+	addWord := func() {
+		if word.Len() == 0 {
+			return
+		}
+
+		addSpace()
+		curWidth += wordLen
+		buf.Write(word.Bytes())
+		word.Reset()
+		wordLen = 0
+	}
+
+	addNewline := func() {
+		buf.WriteByte('\n')
+		curWidth = 0
+		space.Reset()
+	}
+
+	i := 0
+	for i < len(b) {
+		state, action := parser.Table.Transition(pstate, b[i])
+		if state == parser.Utf8State {
+			var width int
+			cluster, _, width, _ = uniseg.FirstGraphemeCluster(b[i:], -1)
+			if m == WcWidth {
+				width = runewidth.StringWidth(string(cluster))
+			}
+			i += len(cluster)
+
+			r, _ := utf8.DecodeRune(cluster)
+			switch {
+			case r != utf8.RuneError && unicode.IsSpace(r) && r != nbsp: // nbsp is a non-breaking space
+				addWord()
+				space.WriteRune(r)
+			case bytes.ContainsAny(cluster, breakpoints):
+				addSpace()
+				if curWidth+wordLen+width > limit {
+					word.Write(cluster)
+					wordLen += width
+				} else {
+					addWord()
+					buf.Write(cluster)
+					curWidth += width
+				}
+			default:
+				if wordLen+width > limit {
+					// Hardwrap the word if it's too long
+					addWord()
+				}
+
+				word.Write(cluster)
+				wordLen += width
+
+				if curWidth+wordLen+space.Len() > limit {
+					addNewline()
+				}
+			}
+
+			pstate = parser.GroundState
+			continue
+		}
+
+		switch action {
+		case parser.PrintAction, parser.ExecuteAction:
+			switch r := rune(b[i]); {
+			case r == '\n':
+				if wordLen == 0 {
+					if curWidth+space.Len() > limit {
+						curWidth = 0
+					} else {
+						// preserve whitespaces
+						buf.Write(space.Bytes())
+					}
+					space.Reset()
+				}
+
+				addWord()
+				addNewline()
+			case unicode.IsSpace(r):
+				addWord()
+				space.WriteRune(r)
+			case r == '-':
+				fallthrough
+			case runeContainsAny(r, breakpoints):
+				addSpace()
+				if curWidth+wordLen >= limit {
+					// We can't fit the breakpoint in the current line, treat
+					// it as part of the word.
+					word.WriteRune(r)
+					wordLen++
+				} else {
+					addWord()
+					buf.WriteRune(r)
+					curWidth++
+				}
+			default:
+				if curWidth == limit {
+					addNewline()
+				}
+				word.WriteRune(r)
+				wordLen++
+
+				if wordLen == limit {
+					// Hardwrap the word if it's too long
+					addWord()
+				}
+
+				if curWidth+wordLen+space.Len() > limit {
+					addNewline()
+				}
+			}
+
+		default:
+			word.WriteByte(b[i])
+		}
+
+		// We manage the UTF8 state separately manually above.
+		if pstate != parser.Utf8State {
+			pstate = state
+		}
+		i++
+	}
+
+	if wordLen == 0 {
+		if curWidth+space.Len() > limit {
+			curWidth = 0
+		} else {
+			// preserve whitespaces
+			buf.Write(space.Bytes())
+		}
+		space.Reset()
+	}
+
+	addWord()
+
+	return buf.String()
+}
+
+func runeContainsAny(r rune, s string) bool {
+	for _, c := range s {
+		if c == r {
+			return true
+		}
+	}
+	return false
+}
diff --git a/vendor/github.com/charmbracelet/x/ansi/xterm.go b/vendor/github.com/charmbracelet/x/ansi/xterm.go
new file mode 100644
index 0000000..83fd4bd
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/ansi/xterm.go
@@ -0,0 +1,138 @@
+package ansi
+
+import "strconv"
+
+// KeyModifierOptions (XTMODKEYS) sets/resets xterm key modifier options.
+//
+// Default is 0.
+//
+//	CSI > Pp m
+//	CSI > Pp ; Pv m
+//
+// If Pv is omitted, the resource is reset to its initial value.
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+func KeyModifierOptions(p int, vs ...int) string {
+	var pp, pv string
+	if p > 0 {
+		pp = strconv.Itoa(p)
+	}
+
+	if len(vs) == 0 {
+		return "\x1b[>" + strconv.Itoa(p) + "m"
+	}
+
+	v := vs[0]
+	if v > 0 {
+		pv = strconv.Itoa(v)
+		return "\x1b[>" + pp + ";" + pv + "m"
+	}
+
+	return "\x1b[>" + pp + "m"
+}
+
+// XTMODKEYS is an alias for [KeyModifierOptions].
+func XTMODKEYS(p int, vs ...int) string {
+	return KeyModifierOptions(p, vs...)
+}
+
+// SetKeyModifierOptions sets xterm key modifier options.
+// This is an alias for [KeyModifierOptions].
+func SetKeyModifierOptions(pp int, pv int) string {
+	return KeyModifierOptions(pp, pv)
+}
+
+// ResetKeyModifierOptions resets xterm key modifier options.
+// This is an alias for [KeyModifierOptions].
+func ResetKeyModifierOptions(pp int) string {
+	return KeyModifierOptions(pp)
+}
+
+// QueryKeyModifierOptions (XTQMODKEYS) requests xterm key modifier options.
+//
+// Default is 0.
+//
+//	CSI ? Pp m
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+func QueryKeyModifierOptions(pp int) string {
+	var p string
+	if pp > 0 {
+		p = strconv.Itoa(pp)
+	}
+	return "\x1b[?" + p + "m"
+}
+
+// XTQMODKEYS is an alias for [QueryKeyModifierOptions].
+func XTQMODKEYS(pp int) string {
+	return QueryKeyModifierOptions(pp)
+}
+
+// Modify Other Keys (modifyOtherKeys) is an xterm feature that allows the
+// terminal to modify the behavior of certain keys to send different escape
+// sequences when pressed.
+//
+// See: https://invisible-island.net/xterm/manpage/xterm.html#VT100-Widget-Resources:modifyOtherKeys
+const (
+	SetModifyOtherKeys1  = "\x1b[>4;1m"
+	SetModifyOtherKeys2  = "\x1b[>4;2m"
+	ResetModifyOtherKeys = "\x1b[>4m"
+	QueryModifyOtherKeys = "\x1b[?4m"
+)
+
+// ModifyOtherKeys returns a sequence that sets XTerm modifyOtherKeys mode.
+// The mode argument specifies the mode to set.
+//
+//	0: Disable modifyOtherKeys mode.
+//	1: Enable modifyOtherKeys mode 1.
+//	2: Enable modifyOtherKeys mode 2.
+//
+//	CSI > 4 ; mode m
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+// See: https://invisible-island.net/xterm/manpage/xterm.html#VT100-Widget-Resources:modifyOtherKeys
+//
+// Deprecated: use [SetModifyOtherKeys1] or [SetModifyOtherKeys2] instead.
+func ModifyOtherKeys(mode int) string {
+	return "\x1b[>4;" + strconv.Itoa(mode) + "m"
+}
+
+// DisableModifyOtherKeys disables the modifyOtherKeys mode.
+//
+//	CSI > 4 ; 0 m
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+// See: https://invisible-island.net/xterm/manpage/xterm.html#VT100-Widget-Resources:modifyOtherKeys
+//
+// Deprecated: use [ResetModifyOtherKeys] instead.
+const DisableModifyOtherKeys = "\x1b[>4;0m"
+
+// EnableModifyOtherKeys1 enables the modifyOtherKeys mode 1.
+//
+//	CSI > 4 ; 1 m
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+// See: https://invisible-island.net/xterm/manpage/xterm.html#VT100-Widget-Resources:modifyOtherKeys
+//
+// Deprecated: use [SetModifyOtherKeys1] instead.
+const EnableModifyOtherKeys1 = "\x1b[>4;1m"
+
+// EnableModifyOtherKeys2 enables the modifyOtherKeys mode 2.
+//
+//	CSI > 4 ; 2 m
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+// See: https://invisible-island.net/xterm/manpage/xterm.html#VT100-Widget-Resources:modifyOtherKeys
+//
+// Deprecated: use [SetModifyOtherKeys2] instead.
+const EnableModifyOtherKeys2 = "\x1b[>4;2m"
+
+// RequestModifyOtherKeys requests the modifyOtherKeys mode.
+//
+//	CSI ? 4  m
+//
+// See: https://invisible-island.net/xterm/ctlseqs/ctlseqs.html#h3-Functions-using-CSI-_-ordered-by-the-final-character_s_
+// See: https://invisible-island.net/xterm/manpage/xterm.html#VT100-Widget-Resources:modifyOtherKeys
+//
+// Deprecated: use [QueryModifyOtherKeys] instead.
+const RequestModifyOtherKeys = "\x1b[?4m"
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/LICENSE b/vendor/github.com/charmbracelet/x/cellbuf/LICENSE
new file mode 100644
index 0000000..65a5654
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2023 Charmbracelet, Inc.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/buffer.go b/vendor/github.com/charmbracelet/x/cellbuf/buffer.go
new file mode 100644
index 0000000..790d1f7
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/buffer.go
@@ -0,0 +1,473 @@
+package cellbuf
+
+import (
+	"strings"
+
+	"github.com/mattn/go-runewidth"
+	"github.com/rivo/uniseg"
+)
+
+// NewCell returns a new cell. This is a convenience function that initializes a
+// new cell with the given content. The cell's width is determined by the
+// content using [runewidth.RuneWidth].
+// This will only account for the first combined rune in the content. If the
+// content is empty, it will return an empty cell with a width of 0.
+func NewCell(r rune, comb ...rune) (c *Cell) {
+	c = new(Cell)
+	c.Rune = r
+	c.Width = runewidth.RuneWidth(r)
+	for _, r := range comb {
+		if runewidth.RuneWidth(r) > 0 {
+			break
+		}
+		c.Comb = append(c.Comb, r)
+	}
+	c.Comb = comb
+	c.Width = runewidth.StringWidth(string(append([]rune{r}, comb...)))
+	return
+}
+
+// NewCellString returns a new cell with the given string content. This is a
+// convenience function that initializes a new cell with the given content. The
+// cell's width is determined by the content using [runewidth.StringWidth].
+// This will only use the first combined rune in the string. If the string is
+// empty, it will return an empty cell with a width of 0.
+func NewCellString(s string) (c *Cell) {
+	c = new(Cell)
+	for i, r := range s {
+		if i == 0 {
+			c.Rune = r
+			// We only care about the first rune's width
+			c.Width = runewidth.RuneWidth(r)
+		} else {
+			if runewidth.RuneWidth(r) > 0 {
+				break
+			}
+			c.Comb = append(c.Comb, r)
+		}
+	}
+	return
+}
+
+// NewGraphemeCell returns a new cell. This is a convenience function that
+// initializes a new cell with the given content. The cell's width is determined
+// by the content using [uniseg.FirstGraphemeClusterInString].
+// This is used when the content is a grapheme cluster i.e. a sequence of runes
+// that form a single visual unit.
+// This will only return the first grapheme cluster in the string. If the
+// string is empty, it will return an empty cell with a width of 0.
+func NewGraphemeCell(s string) (c *Cell) {
+	g, _, w, _ := uniseg.FirstGraphemeClusterInString(s, -1)
+	return newGraphemeCell(g, w)
+}
+
+func newGraphemeCell(s string, w int) (c *Cell) {
+	c = new(Cell)
+	c.Width = w
+	for i, r := range s {
+		if i == 0 {
+			c.Rune = r
+		} else {
+			c.Comb = append(c.Comb, r)
+		}
+	}
+	return
+}
+
+// Line represents a line in the terminal.
+// A nil cell represents an blank cell, a cell with a space character and a
+// width of 1.
+// If a cell has no content and a width of 0, it is a placeholder for a wide
+// cell.
+type Line []*Cell
+
+// Width returns the width of the line.
+func (l Line) Width() int {
+	return len(l)
+}
+
+// Len returns the length of the line.
+func (l Line) Len() int {
+	return len(l)
+}
+
+// String returns the string representation of the line. Any trailing spaces
+// are removed.
+func (l Line) String() (s string) {
+	for _, c := range l {
+		if c == nil {
+			s += " "
+		} else if c.Empty() {
+			continue
+		} else {
+			s += c.String()
+		}
+	}
+	s = strings.TrimRight(s, " ")
+	return
+}
+
+// At returns the cell at the given x position.
+// If the cell does not exist, it returns nil.
+func (l Line) At(x int) *Cell {
+	if x < 0 || x >= len(l) {
+		return nil
+	}
+
+	c := l[x]
+	if c == nil {
+		newCell := BlankCell
+		return &newCell
+	}
+
+	return c
+}
+
+// Set sets the cell at the given x position. If a wide cell is given, it will
+// set the cell and the following cells to [EmptyCell]. It returns true if the
+// cell was set.
+func (l Line) Set(x int, c *Cell) bool {
+	return l.set(x, c, true)
+}
+
+func (l Line) set(x int, c *Cell, clone bool) bool {
+	width := l.Width()
+	if x < 0 || x >= width {
+		return false
+	}
+
+	// When a wide cell is partially overwritten, we need
+	// to fill the rest of the cell with space cells to
+	// avoid rendering issues.
+	prev := l.At(x)
+	if prev != nil && prev.Width > 1 {
+		// Writing to the first wide cell
+		for j := 0; j < prev.Width && x+j < l.Width(); j++ {
+			l[x+j] = prev.Clone().Blank()
+		}
+	} else if prev != nil && prev.Width == 0 {
+		// Writing to wide cell placeholders
+		for j := 1; j < maxCellWidth && x-j >= 0; j++ {
+			wide := l.At(x - j)
+			if wide != nil && wide.Width > 1 && j < wide.Width {
+				for k := 0; k < wide.Width; k++ {
+					l[x-j+k] = wide.Clone().Blank()
+				}
+				break
+			}
+		}
+	}
+
+	if clone && c != nil {
+		// Clone the cell if not nil.
+		c = c.Clone()
+	}
+
+	if c != nil && x+c.Width > width {
+		// If the cell is too wide, we write blanks with the same style.
+		for i := 0; i < c.Width && x+i < width; i++ {
+			l[x+i] = c.Clone().Blank()
+		}
+	} else {
+		l[x] = c
+
+		// Mark wide cells with an empty cell zero width
+		// We set the wide cell down below
+		if c != nil && c.Width > 1 {
+			for j := 1; j < c.Width && x+j < l.Width(); j++ {
+				var wide Cell
+				l[x+j] = &wide
+			}
+		}
+	}
+
+	return true
+}
+
+// Buffer is a 2D grid of cells representing a screen or terminal.
+type Buffer struct {
+	// Lines holds the lines of the buffer.
+	Lines []Line
+}
+
+// NewBuffer creates a new buffer with the given width and height.
+// This is a convenience function that initializes a new buffer and resizes it.
+func NewBuffer(width int, height int) *Buffer {
+	b := new(Buffer)
+	b.Resize(width, height)
+	return b
+}
+
+// String returns the string representation of the buffer.
+func (b *Buffer) String() (s string) {
+	for i, l := range b.Lines {
+		s += l.String()
+		if i < len(b.Lines)-1 {
+			s += "\r\n"
+		}
+	}
+	return
+}
+
+// Line returns a pointer to the line at the given y position.
+// If the line does not exist, it returns nil.
+func (b *Buffer) Line(y int) Line {
+	if y < 0 || y >= len(b.Lines) {
+		return nil
+	}
+	return b.Lines[y]
+}
+
+// Cell implements Screen.
+func (b *Buffer) Cell(x int, y int) *Cell {
+	if y < 0 || y >= len(b.Lines) {
+		return nil
+	}
+	return b.Lines[y].At(x)
+}
+
+// maxCellWidth is the maximum width a terminal cell can get.
+const maxCellWidth = 4
+
+// SetCell sets the cell at the given x, y position.
+func (b *Buffer) SetCell(x, y int, c *Cell) bool {
+	return b.setCell(x, y, c, true)
+}
+
+// setCell sets the cell at the given x, y position. This will always clone and
+// allocates a new cell if c is not nil.
+func (b *Buffer) setCell(x, y int, c *Cell, clone bool) bool {
+	if y < 0 || y >= len(b.Lines) {
+		return false
+	}
+	return b.Lines[y].set(x, c, clone)
+}
+
+// Height implements Screen.
+func (b *Buffer) Height() int {
+	return len(b.Lines)
+}
+
+// Width implements Screen.
+func (b *Buffer) Width() int {
+	if len(b.Lines) == 0 {
+		return 0
+	}
+	return b.Lines[0].Width()
+}
+
+// Bounds returns the bounds of the buffer.
+func (b *Buffer) Bounds() Rectangle {
+	return Rect(0, 0, b.Width(), b.Height())
+}
+
+// Resize resizes the buffer to the given width and height.
+func (b *Buffer) Resize(width int, height int) {
+	if width == 0 || height == 0 {
+		b.Lines = nil
+		return
+	}
+
+	if width > b.Width() {
+		line := make(Line, width-b.Width())
+		for i := range b.Lines {
+			b.Lines[i] = append(b.Lines[i], line...)
+		}
+	} else if width < b.Width() {
+		for i := range b.Lines {
+			b.Lines[i] = b.Lines[i][:width]
+		}
+	}
+
+	if height > len(b.Lines) {
+		for i := len(b.Lines); i < height; i++ {
+			b.Lines = append(b.Lines, make(Line, width))
+		}
+	} else if height < len(b.Lines) {
+		b.Lines = b.Lines[:height]
+	}
+}
+
+// FillRect fills the buffer with the given cell and rectangle.
+func (b *Buffer) FillRect(c *Cell, rect Rectangle) {
+	cellWidth := 1
+	if c != nil && c.Width > 1 {
+		cellWidth = c.Width
+	}
+	for y := rect.Min.Y; y < rect.Max.Y; y++ {
+		for x := rect.Min.X; x < rect.Max.X; x += cellWidth {
+			b.setCell(x, y, c, false) //nolint:errcheck
+		}
+	}
+}
+
+// Fill fills the buffer with the given cell and rectangle.
+func (b *Buffer) Fill(c *Cell) {
+	b.FillRect(c, b.Bounds())
+}
+
+// Clear clears the buffer with space cells and rectangle.
+func (b *Buffer) Clear() {
+	b.ClearRect(b.Bounds())
+}
+
+// ClearRect clears the buffer with space cells within the specified
+// rectangles. Only cells within the rectangle's bounds are affected.
+func (b *Buffer) ClearRect(rect Rectangle) {
+	b.FillRect(nil, rect)
+}
+
+// InsertLine inserts n lines at the given line position, with the given
+// optional cell, within the specified rectangles. If no rectangles are
+// specified, it inserts lines in the entire buffer. Only cells within the
+// rectangle's horizontal bounds are affected. Lines are pushed out of the
+// rectangle bounds and lost. This follows terminal [ansi.IL] behavior.
+// It returns the pushed out lines.
+func (b *Buffer) InsertLine(y, n int, c *Cell) {
+	b.InsertLineRect(y, n, c, b.Bounds())
+}
+
+// InsertLineRect inserts new lines at the given line position, with the
+// given optional cell, within the rectangle bounds. Only cells within the
+// rectangle's horizontal bounds are affected. Lines are pushed out of the
+// rectangle bounds and lost. This follows terminal [ansi.IL] behavior.
+func (b *Buffer) InsertLineRect(y, n int, c *Cell, rect Rectangle) {
+	if n <= 0 || y < rect.Min.Y || y >= rect.Max.Y || y >= b.Height() {
+		return
+	}
+
+	// Limit number of lines to insert to available space
+	if y+n > rect.Max.Y {
+		n = rect.Max.Y - y
+	}
+
+	// Move existing lines down within the bounds
+	for i := rect.Max.Y - 1; i >= y+n; i-- {
+		for x := rect.Min.X; x < rect.Max.X; x++ {
+			// We don't need to clone c here because we're just moving lines down.
+			b.setCell(x, i, b.Lines[i-n][x], false)
+		}
+	}
+
+	// Clear the newly inserted lines within bounds
+	for i := y; i < y+n; i++ {
+		for x := rect.Min.X; x < rect.Max.X; x++ {
+			b.setCell(x, i, c, true)
+		}
+	}
+}
+
+// DeleteLineRect deletes lines at the given line position, with the given
+// optional cell, within the rectangle bounds. Only cells within the
+// rectangle's bounds are affected. Lines are shifted up within the bounds and
+// new blank lines are created at the bottom. This follows terminal [ansi.DL]
+// behavior.
+func (b *Buffer) DeleteLineRect(y, n int, c *Cell, rect Rectangle) {
+	if n <= 0 || y < rect.Min.Y || y >= rect.Max.Y || y >= b.Height() {
+		return
+	}
+
+	// Limit deletion count to available space in scroll region
+	if n > rect.Max.Y-y {
+		n = rect.Max.Y - y
+	}
+
+	// Shift cells up within the bounds
+	for dst := y; dst < rect.Max.Y-n; dst++ {
+		src := dst + n
+		for x := rect.Min.X; x < rect.Max.X; x++ {
+			// We don't need to clone c here because we're just moving cells up.
+			// b.lines[dst][x] = b.lines[src][x]
+			b.setCell(x, dst, b.Lines[src][x], false)
+		}
+	}
+
+	// Fill the bottom n lines with blank cells
+	for i := rect.Max.Y - n; i < rect.Max.Y; i++ {
+		for x := rect.Min.X; x < rect.Max.X; x++ {
+			b.setCell(x, i, c, true)
+		}
+	}
+}
+
+// DeleteLine deletes n lines at the given line position, with the given
+// optional cell, within the specified rectangles. If no rectangles are
+// specified, it deletes lines in the entire buffer.
+func (b *Buffer) DeleteLine(y, n int, c *Cell) {
+	b.DeleteLineRect(y, n, c, b.Bounds())
+}
+
+// InsertCell inserts new cells at the given position, with the given optional
+// cell, within the specified rectangles. If no rectangles are specified, it
+// inserts cells in the entire buffer. This follows terminal [ansi.ICH]
+// behavior.
+func (b *Buffer) InsertCell(x, y, n int, c *Cell) {
+	b.InsertCellRect(x, y, n, c, b.Bounds())
+}
+
+// InsertCellRect inserts new cells at the given position, with the given
+// optional cell, within the rectangle bounds. Only cells within the
+// rectangle's bounds are affected, following terminal [ansi.ICH] behavior.
+func (b *Buffer) InsertCellRect(x, y, n int, c *Cell, rect Rectangle) {
+	if n <= 0 || y < rect.Min.Y || y >= rect.Max.Y || y >= b.Height() ||
+		x < rect.Min.X || x >= rect.Max.X || x >= b.Width() {
+		return
+	}
+
+	// Limit number of cells to insert to available space
+	if x+n > rect.Max.X {
+		n = rect.Max.X - x
+	}
+
+	// Move existing cells within rectangle bounds to the right
+	for i := rect.Max.X - 1; i >= x+n && i-n >= rect.Min.X; i-- {
+		// We don't need to clone c here because we're just moving cells to the
+		// right.
+		// b.lines[y][i] = b.lines[y][i-n]
+		b.setCell(i, y, b.Lines[y][i-n], false)
+	}
+
+	// Clear the newly inserted cells within rectangle bounds
+	for i := x; i < x+n && i < rect.Max.X; i++ {
+		b.setCell(i, y, c, true)
+	}
+}
+
+// DeleteCell deletes cells at the given position, with the given optional
+// cell, within the specified rectangles. If no rectangles are specified, it
+// deletes cells in the entire buffer. This follows terminal [ansi.DCH]
+// behavior.
+func (b *Buffer) DeleteCell(x, y, n int, c *Cell) {
+	b.DeleteCellRect(x, y, n, c, b.Bounds())
+}
+
+// DeleteCellRect deletes cells at the given position, with the given
+// optional cell, within the rectangle bounds. Only cells within the
+// rectangle's bounds are affected, following terminal [ansi.DCH] behavior.
+func (b *Buffer) DeleteCellRect(x, y, n int, c *Cell, rect Rectangle) {
+	if n <= 0 || y < rect.Min.Y || y >= rect.Max.Y || y >= b.Height() ||
+		x < rect.Min.X || x >= rect.Max.X || x >= b.Width() {
+		return
+	}
+
+	// Calculate how many positions we can actually delete
+	remainingCells := rect.Max.X - x
+	if n > remainingCells {
+		n = remainingCells
+	}
+
+	// Shift the remaining cells to the left
+	for i := x; i < rect.Max.X-n; i++ {
+		if i+n < rect.Max.X {
+			// We don't need to clone c here because we're just moving cells to
+			// the left.
+			// b.lines[y][i] = b.lines[y][i+n]
+			b.setCell(i, y, b.Lines[y][i+n], false)
+		}
+	}
+
+	// Fill the vacated positions with the given cell
+	for i := rect.Max.X - n; i < rect.Max.X; i++ {
+		b.setCell(i, y, c, true)
+	}
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/cell.go b/vendor/github.com/charmbracelet/x/cellbuf/cell.go
new file mode 100644
index 0000000..991c919
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/cell.go
@@ -0,0 +1,503 @@
+package cellbuf
+
+import (
+	"github.com/charmbracelet/x/ansi"
+)
+
+var (
+	// BlankCell is a cell with a single space, width of 1, and no style or link.
+	BlankCell = Cell{Rune: ' ', Width: 1}
+
+	// EmptyCell is just an empty cell used for comparisons and as a placeholder
+	// for wide cells.
+	EmptyCell = Cell{}
+)
+
+// Cell represents a single cell in the terminal screen.
+type Cell struct {
+	// The style of the cell. Nil style means no style. Zero value prints a
+	// reset sequence.
+	Style Style
+
+	// Link is the hyperlink of the cell.
+	Link Link
+
+	// Comb is the combining runes of the cell. This is nil if the cell is a
+	// single rune or if it's a zero width cell that is part of a wider cell.
+	Comb []rune
+
+	// Width is the mono-space width of the grapheme cluster.
+	Width int
+
+	// Rune is the main rune of the cell. This is zero if the cell is part of a
+	// wider cell.
+	Rune rune
+}
+
+// Append appends runes to the cell without changing the width. This is useful
+// when we want to use the cell to store escape sequences or other runes that
+// don't affect the width of the cell.
+func (c *Cell) Append(r ...rune) {
+	for i, r := range r {
+		if i == 0 && c.Rune == 0 {
+			c.Rune = r
+			continue
+		}
+		c.Comb = append(c.Comb, r)
+	}
+}
+
+// String returns the string content of the cell excluding any styles, links,
+// and escape sequences.
+func (c Cell) String() string {
+	if c.Rune == 0 {
+		return ""
+	}
+	if len(c.Comb) == 0 {
+		return string(c.Rune)
+	}
+	return string(append([]rune{c.Rune}, c.Comb...))
+}
+
+// Equal returns whether the cell is equal to the other cell.
+func (c *Cell) Equal(o *Cell) bool {
+	return o != nil &&
+		c.Width == o.Width &&
+		c.Rune == o.Rune &&
+		runesEqual(c.Comb, o.Comb) &&
+		c.Style.Equal(&o.Style) &&
+		c.Link.Equal(&o.Link)
+}
+
+// Empty returns whether the cell is an empty cell. An empty cell is a cell
+// with a width of 0, a rune of 0, and no combining runes.
+func (c Cell) Empty() bool {
+	return c.Width == 0 &&
+		c.Rune == 0 &&
+		len(c.Comb) == 0
+}
+
+// Reset resets the cell to the default state zero value.
+func (c *Cell) Reset() {
+	c.Rune = 0
+	c.Comb = nil
+	c.Width = 0
+	c.Style.Reset()
+	c.Link.Reset()
+}
+
+// Clear returns whether the cell consists of only attributes that don't
+// affect appearance of a space character.
+func (c *Cell) Clear() bool {
+	return c.Rune == ' ' && len(c.Comb) == 0 && c.Width == 1 && c.Style.Clear() && c.Link.Empty()
+}
+
+// Clone returns a copy of the cell.
+func (c *Cell) Clone() (n *Cell) {
+	n = new(Cell)
+	*n = *c
+	return
+}
+
+// Blank makes the cell a blank cell by setting the rune to a space, comb to
+// nil, and the width to 1.
+func (c *Cell) Blank() *Cell {
+	c.Rune = ' '
+	c.Comb = nil
+	c.Width = 1
+	return c
+}
+
+// Link represents a hyperlink in the terminal screen.
+type Link struct {
+	URL    string
+	Params string
+}
+
+// String returns a string representation of the hyperlink.
+func (h Link) String() string {
+	return h.URL
+}
+
+// Reset resets the hyperlink to the default state zero value.
+func (h *Link) Reset() {
+	h.URL = ""
+	h.Params = ""
+}
+
+// Equal returns whether the hyperlink is equal to the other hyperlink.
+func (h *Link) Equal(o *Link) bool {
+	return o != nil && h.URL == o.URL && h.Params == o.Params
+}
+
+// Empty returns whether the hyperlink is empty.
+func (h Link) Empty() bool {
+	return h.URL == "" && h.Params == ""
+}
+
+// AttrMask is a bitmask for text attributes that can change the look of text.
+// These attributes can be combined to create different styles.
+type AttrMask uint8
+
+// These are the available text attributes that can be combined to create
+// different styles.
+const (
+	BoldAttr AttrMask = 1 << iota
+	FaintAttr
+	ItalicAttr
+	SlowBlinkAttr
+	RapidBlinkAttr
+	ReverseAttr
+	ConcealAttr
+	StrikethroughAttr
+
+	ResetAttr AttrMask = 0
+)
+
+// UnderlineStyle is the style of underline to use for text.
+type UnderlineStyle = ansi.UnderlineStyle
+
+// These are the available underline styles.
+const (
+	NoUnderline     = ansi.NoUnderlineStyle
+	SingleUnderline = ansi.SingleUnderlineStyle
+	DoubleUnderline = ansi.DoubleUnderlineStyle
+	CurlyUnderline  = ansi.CurlyUnderlineStyle
+	DottedUnderline = ansi.DottedUnderlineStyle
+	DashedUnderline = ansi.DashedUnderlineStyle
+)
+
+// Style represents the Style of a cell.
+type Style struct {
+	Fg      ansi.Color
+	Bg      ansi.Color
+	Ul      ansi.Color
+	Attrs   AttrMask
+	UlStyle UnderlineStyle
+}
+
+// Sequence returns the ANSI sequence that sets the style.
+func (s Style) Sequence() string {
+	if s.Empty() {
+		return ansi.ResetStyle
+	}
+
+	var b ansi.Style
+
+	if s.Attrs != 0 {
+		if s.Attrs&BoldAttr != 0 {
+			b = b.Bold()
+		}
+		if s.Attrs&FaintAttr != 0 {
+			b = b.Faint()
+		}
+		if s.Attrs&ItalicAttr != 0 {
+			b = b.Italic()
+		}
+		if s.Attrs&SlowBlinkAttr != 0 {
+			b = b.SlowBlink()
+		}
+		if s.Attrs&RapidBlinkAttr != 0 {
+			b = b.RapidBlink()
+		}
+		if s.Attrs&ReverseAttr != 0 {
+			b = b.Reverse()
+		}
+		if s.Attrs&ConcealAttr != 0 {
+			b = b.Conceal()
+		}
+		if s.Attrs&StrikethroughAttr != 0 {
+			b = b.Strikethrough()
+		}
+	}
+	if s.UlStyle != NoUnderline {
+		switch s.UlStyle {
+		case SingleUnderline:
+			b = b.Underline()
+		case DoubleUnderline:
+			b = b.DoubleUnderline()
+		case CurlyUnderline:
+			b = b.CurlyUnderline()
+		case DottedUnderline:
+			b = b.DottedUnderline()
+		case DashedUnderline:
+			b = b.DashedUnderline()
+		}
+	}
+	if s.Fg != nil {
+		b = b.ForegroundColor(s.Fg)
+	}
+	if s.Bg != nil {
+		b = b.BackgroundColor(s.Bg)
+	}
+	if s.Ul != nil {
+		b = b.UnderlineColor(s.Ul)
+	}
+
+	return b.String()
+}
+
+// DiffSequence returns the ANSI sequence that sets the style as a diff from
+// another style.
+func (s Style) DiffSequence(o Style) string {
+	if o.Empty() {
+		return s.Sequence()
+	}
+
+	var b ansi.Style
+
+	if !colorEqual(s.Fg, o.Fg) {
+		b = b.ForegroundColor(s.Fg)
+	}
+
+	if !colorEqual(s.Bg, o.Bg) {
+		b = b.BackgroundColor(s.Bg)
+	}
+
+	if !colorEqual(s.Ul, o.Ul) {
+		b = b.UnderlineColor(s.Ul)
+	}
+
+	var (
+		noBlink  bool
+		isNormal bool
+	)
+
+	if s.Attrs != o.Attrs {
+		if s.Attrs&BoldAttr != o.Attrs&BoldAttr {
+			if s.Attrs&BoldAttr != 0 {
+				b = b.Bold()
+			} else if !isNormal {
+				isNormal = true
+				b = b.NormalIntensity()
+			}
+		}
+		if s.Attrs&FaintAttr != o.Attrs&FaintAttr {
+			if s.Attrs&FaintAttr != 0 {
+				b = b.Faint()
+			} else if !isNormal {
+				b = b.NormalIntensity()
+			}
+		}
+		if s.Attrs&ItalicAttr != o.Attrs&ItalicAttr {
+			if s.Attrs&ItalicAttr != 0 {
+				b = b.Italic()
+			} else {
+				b = b.NoItalic()
+			}
+		}
+		if s.Attrs&SlowBlinkAttr != o.Attrs&SlowBlinkAttr {
+			if s.Attrs&SlowBlinkAttr != 0 {
+				b = b.SlowBlink()
+			} else if !noBlink {
+				noBlink = true
+				b = b.NoBlink()
+			}
+		}
+		if s.Attrs&RapidBlinkAttr != o.Attrs&RapidBlinkAttr {
+			if s.Attrs&RapidBlinkAttr != 0 {
+				b = b.RapidBlink()
+			} else if !noBlink {
+				b = b.NoBlink()
+			}
+		}
+		if s.Attrs&ReverseAttr != o.Attrs&ReverseAttr {
+			if s.Attrs&ReverseAttr != 0 {
+				b = b.Reverse()
+			} else {
+				b = b.NoReverse()
+			}
+		}
+		if s.Attrs&ConcealAttr != o.Attrs&ConcealAttr {
+			if s.Attrs&ConcealAttr != 0 {
+				b = b.Conceal()
+			} else {
+				b = b.NoConceal()
+			}
+		}
+		if s.Attrs&StrikethroughAttr != o.Attrs&StrikethroughAttr {
+			if s.Attrs&StrikethroughAttr != 0 {
+				b = b.Strikethrough()
+			} else {
+				b = b.NoStrikethrough()
+			}
+		}
+	}
+
+	if s.UlStyle != o.UlStyle {
+		b = b.UnderlineStyle(s.UlStyle)
+	}
+
+	return b.String()
+}
+
+// Equal returns true if the style is equal to the other style.
+func (s *Style) Equal(o *Style) bool {
+	return s.Attrs == o.Attrs &&
+		s.UlStyle == o.UlStyle &&
+		colorEqual(s.Fg, o.Fg) &&
+		colorEqual(s.Bg, o.Bg) &&
+		colorEqual(s.Ul, o.Ul)
+}
+
+func colorEqual(c, o ansi.Color) bool {
+	if c == nil && o == nil {
+		return true
+	}
+	if c == nil || o == nil {
+		return false
+	}
+	cr, cg, cb, ca := c.RGBA()
+	or, og, ob, oa := o.RGBA()
+	return cr == or && cg == og && cb == ob && ca == oa
+}
+
+// Bold sets the bold attribute.
+func (s *Style) Bold(v bool) *Style {
+	if v {
+		s.Attrs |= BoldAttr
+	} else {
+		s.Attrs &^= BoldAttr
+	}
+	return s
+}
+
+// Faint sets the faint attribute.
+func (s *Style) Faint(v bool) *Style {
+	if v {
+		s.Attrs |= FaintAttr
+	} else {
+		s.Attrs &^= FaintAttr
+	}
+	return s
+}
+
+// Italic sets the italic attribute.
+func (s *Style) Italic(v bool) *Style {
+	if v {
+		s.Attrs |= ItalicAttr
+	} else {
+		s.Attrs &^= ItalicAttr
+	}
+	return s
+}
+
+// SlowBlink sets the slow blink attribute.
+func (s *Style) SlowBlink(v bool) *Style {
+	if v {
+		s.Attrs |= SlowBlinkAttr
+	} else {
+		s.Attrs &^= SlowBlinkAttr
+	}
+	return s
+}
+
+// RapidBlink sets the rapid blink attribute.
+func (s *Style) RapidBlink(v bool) *Style {
+	if v {
+		s.Attrs |= RapidBlinkAttr
+	} else {
+		s.Attrs &^= RapidBlinkAttr
+	}
+	return s
+}
+
+// Reverse sets the reverse attribute.
+func (s *Style) Reverse(v bool) *Style {
+	if v {
+		s.Attrs |= ReverseAttr
+	} else {
+		s.Attrs &^= ReverseAttr
+	}
+	return s
+}
+
+// Conceal sets the conceal attribute.
+func (s *Style) Conceal(v bool) *Style {
+	if v {
+		s.Attrs |= ConcealAttr
+	} else {
+		s.Attrs &^= ConcealAttr
+	}
+	return s
+}
+
+// Strikethrough sets the strikethrough attribute.
+func (s *Style) Strikethrough(v bool) *Style {
+	if v {
+		s.Attrs |= StrikethroughAttr
+	} else {
+		s.Attrs &^= StrikethroughAttr
+	}
+	return s
+}
+
+// UnderlineStyle sets the underline style.
+func (s *Style) UnderlineStyle(style UnderlineStyle) *Style {
+	s.UlStyle = style
+	return s
+}
+
+// Underline sets the underline attribute.
+// This is a syntactic sugar for [UnderlineStyle].
+func (s *Style) Underline(v bool) *Style {
+	if v {
+		return s.UnderlineStyle(SingleUnderline)
+	}
+	return s.UnderlineStyle(NoUnderline)
+}
+
+// Foreground sets the foreground color.
+func (s *Style) Foreground(c ansi.Color) *Style {
+	s.Fg = c
+	return s
+}
+
+// Background sets the background color.
+func (s *Style) Background(c ansi.Color) *Style {
+	s.Bg = c
+	return s
+}
+
+// UnderlineColor sets the underline color.
+func (s *Style) UnderlineColor(c ansi.Color) *Style {
+	s.Ul = c
+	return s
+}
+
+// Reset resets the style to default.
+func (s *Style) Reset() *Style {
+	s.Fg = nil
+	s.Bg = nil
+	s.Ul = nil
+	s.Attrs = ResetAttr
+	s.UlStyle = NoUnderline
+	return s
+}
+
+// Empty returns true if the style is empty.
+func (s *Style) Empty() bool {
+	return s.Fg == nil && s.Bg == nil && s.Ul == nil && s.Attrs == ResetAttr && s.UlStyle == NoUnderline
+}
+
+// Clear returns whether the style consists of only attributes that don't
+// affect appearance of a space character.
+func (s *Style) Clear() bool {
+	return s.UlStyle == NoUnderline &&
+		s.Attrs&^(BoldAttr|FaintAttr|ItalicAttr|SlowBlinkAttr|RapidBlinkAttr) == 0 &&
+		s.Fg == nil &&
+		s.Bg == nil &&
+		s.Ul == nil
+}
+
+func runesEqual(a, b []rune) bool {
+	if len(a) != len(b) {
+		return false
+	}
+	for i, r := range a {
+		if r != b[i] {
+			return false
+		}
+	}
+	return true
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/errors.go b/vendor/github.com/charmbracelet/x/cellbuf/errors.go
new file mode 100644
index 0000000..64258fe
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/errors.go
@@ -0,0 +1,6 @@
+package cellbuf
+
+import "errors"
+
+// ErrOutOfBounds is returned when the given x, y position is out of bounds.
+var ErrOutOfBounds = errors.New("out of bounds")
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/geom.go b/vendor/github.com/charmbracelet/x/cellbuf/geom.go
new file mode 100644
index 0000000..c12e6fb
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/geom.go
@@ -0,0 +1,21 @@
+package cellbuf
+
+import (
+	"image"
+)
+
+// Position represents an x, y position.
+type Position = image.Point
+
+// Pos is a shorthand for Position{X: x, Y: y}.
+func Pos(x, y int) Position {
+	return image.Pt(x, y)
+}
+
+// Rectange represents a rectangle.
+type Rectangle = image.Rectangle
+
+// Rect is a shorthand for Rectangle.
+func Rect(x, y, w, h int) Rectangle {
+	return image.Rect(x, y, x+w, y+h)
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/hardscroll.go b/vendor/github.com/charmbracelet/x/cellbuf/hardscroll.go
new file mode 100644
index 0000000..402ac06
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/hardscroll.go
@@ -0,0 +1,272 @@
+package cellbuf
+
+import (
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// scrollOptimize optimizes the screen to transform the old buffer into the new
+// buffer.
+func (s *Screen) scrollOptimize() {
+	height := s.newbuf.Height()
+	if s.oldnum == nil || len(s.oldnum) < height {
+		s.oldnum = make([]int, height)
+	}
+
+	// Calculate the indices
+	s.updateHashmap()
+	if len(s.hashtab) < height {
+		return
+	}
+
+	// Pass 1 - from top to bottom scrolling up
+	for i := 0; i < height; {
+		for i < height && (s.oldnum[i] == newIndex || s.oldnum[i] <= i) {
+			i++
+		}
+		if i >= height {
+			break
+		}
+
+		shift := s.oldnum[i] - i // shift > 0
+		start := i
+
+		i++
+		for i < height && s.oldnum[i] != newIndex && s.oldnum[i]-i == shift {
+			i++
+		}
+		end := i - 1 + shift
+
+		if !s.scrolln(shift, start, end, height-1) {
+			continue
+		}
+	}
+
+	// Pass 2 - from bottom to top scrolling down
+	for i := height - 1; i >= 0; {
+		for i >= 0 && (s.oldnum[i] == newIndex || s.oldnum[i] >= i) {
+			i--
+		}
+		if i < 0 {
+			break
+		}
+
+		shift := s.oldnum[i] - i // shift < 0
+		end := i
+
+		i--
+		for i >= 0 && s.oldnum[i] != newIndex && s.oldnum[i]-i == shift {
+			i--
+		}
+
+		start := i + 1 - (-shift)
+		if !s.scrolln(shift, start, end, height-1) {
+			continue
+		}
+	}
+}
+
+// scrolln scrolls the screen up by n lines.
+func (s *Screen) scrolln(n, top, bot, maxY int) (v bool) { //nolint:unparam
+	const (
+		nonDestScrollRegion = false
+		memoryBelow         = false
+	)
+
+	blank := s.clearBlank()
+	if n > 0 {
+		// Scroll up (forward)
+		v = s.scrollUp(n, top, bot, 0, maxY, blank)
+		if !v {
+			s.buf.WriteString(ansi.SetTopBottomMargins(top+1, bot+1))
+
+			// XXX: How should we handle this in inline mode when not using alternate screen?
+			s.cur.X, s.cur.Y = -1, -1
+			v = s.scrollUp(n, top, bot, top, bot, blank)
+			s.buf.WriteString(ansi.SetTopBottomMargins(1, maxY+1))
+			s.cur.X, s.cur.Y = -1, -1
+		}
+
+		if !v {
+			v = s.scrollIdl(n, top, bot-n+1, blank)
+		}
+
+		// Clear newly shifted-in lines.
+		if v &&
+			(nonDestScrollRegion || (memoryBelow && bot == maxY)) {
+			if bot == maxY {
+				s.move(0, bot-n+1)
+				s.clearToBottom(nil)
+			} else {
+				for i := 0; i < n; i++ {
+					s.move(0, bot-i)
+					s.clearToEnd(nil, false)
+				}
+			}
+		}
+	} else if n < 0 {
+		// Scroll down (backward)
+		v = s.scrollDown(-n, top, bot, 0, maxY, blank)
+		if !v {
+			s.buf.WriteString(ansi.SetTopBottomMargins(top+1, bot+1))
+
+			// XXX: How should we handle this in inline mode when not using alternate screen?
+			s.cur.X, s.cur.Y = -1, -1
+			v = s.scrollDown(-n, top, bot, top, bot, blank)
+			s.buf.WriteString(ansi.SetTopBottomMargins(1, maxY+1))
+			s.cur.X, s.cur.Y = -1, -1
+
+			if !v {
+				v = s.scrollIdl(-n, bot+n+1, top, blank)
+			}
+
+			// Clear newly shifted-in lines.
+			if v &&
+				(nonDestScrollRegion || (memoryBelow && top == 0)) {
+				for i := 0; i < -n; i++ {
+					s.move(0, top+i)
+					s.clearToEnd(nil, false)
+				}
+			}
+		}
+	}
+
+	if !v {
+		return
+	}
+
+	s.scrollBuffer(s.curbuf, n, top, bot, blank)
+
+	// shift hash values too, they can be reused
+	s.scrollOldhash(n, top, bot)
+
+	return true
+}
+
+// scrollBuffer scrolls the buffer by n lines.
+func (s *Screen) scrollBuffer(b *Buffer, n, top, bot int, blank *Cell) {
+	if top < 0 || bot < top || bot >= b.Height() {
+		// Nothing to scroll
+		return
+	}
+
+	if n < 0 {
+		// shift n lines downwards
+		limit := top - n
+		for line := bot; line >= limit && line >= 0 && line >= top; line-- {
+			copy(b.Lines[line], b.Lines[line+n])
+		}
+		for line := top; line < limit && line <= b.Height()-1 && line <= bot; line++ {
+			b.FillRect(blank, Rect(0, line, b.Width(), 1))
+		}
+	}
+
+	if n > 0 {
+		// shift n lines upwards
+		limit := bot - n
+		for line := top; line <= limit && line <= b.Height()-1 && line <= bot; line++ {
+			copy(b.Lines[line], b.Lines[line+n])
+		}
+		for line := bot; line > limit && line >= 0 && line >= top; line-- {
+			b.FillRect(blank, Rect(0, line, b.Width(), 1))
+		}
+	}
+
+	s.touchLine(b.Width(), b.Height(), top, bot-top+1, true)
+}
+
+// touchLine marks the line as touched.
+func (s *Screen) touchLine(width, height, y, n int, changed bool) {
+	if n < 0 || y < 0 || y >= height {
+		return // Nothing to touch
+	}
+
+	for i := y; i < y+n && i < height; i++ {
+		if changed {
+			s.touch[i] = lineData{firstCell: 0, lastCell: width - 1}
+		} else {
+			delete(s.touch, i)
+		}
+	}
+}
+
+// scrollUp scrolls the screen up by n lines.
+func (s *Screen) scrollUp(n, top, bot, minY, maxY int, blank *Cell) bool {
+	if n == 1 && top == minY && bot == maxY {
+		s.move(0, bot)
+		s.updatePen(blank)
+		s.buf.WriteByte('\n')
+	} else if n == 1 && bot == maxY {
+		s.move(0, top)
+		s.updatePen(blank)
+		s.buf.WriteString(ansi.DeleteLine(1))
+	} else if top == minY && bot == maxY {
+		if s.xtermLike {
+			s.move(0, bot)
+		} else {
+			s.move(0, top)
+		}
+		s.updatePen(blank)
+		if s.xtermLike {
+			s.buf.WriteString(ansi.ScrollUp(n))
+		} else {
+			s.buf.WriteString(strings.Repeat("\n", n))
+		}
+	} else if bot == maxY {
+		s.move(0, top)
+		s.updatePen(blank)
+		s.buf.WriteString(ansi.DeleteLine(n))
+	} else {
+		return false
+	}
+	return true
+}
+
+// scrollDown scrolls the screen down by n lines.
+func (s *Screen) scrollDown(n, top, bot, minY, maxY int, blank *Cell) bool {
+	if n == 1 && top == minY && bot == maxY {
+		s.move(0, top)
+		s.updatePen(blank)
+		s.buf.WriteString(ansi.ReverseIndex)
+	} else if n == 1 && bot == maxY {
+		s.move(0, top)
+		s.updatePen(blank)
+		s.buf.WriteString(ansi.InsertLine(1))
+	} else if top == minY && bot == maxY {
+		s.move(0, top)
+		s.updatePen(blank)
+		if s.xtermLike {
+			s.buf.WriteString(ansi.ScrollDown(n))
+		} else {
+			s.buf.WriteString(strings.Repeat(ansi.ReverseIndex, n))
+		}
+	} else if bot == maxY {
+		s.move(0, top)
+		s.updatePen(blank)
+		s.buf.WriteString(ansi.InsertLine(n))
+	} else {
+		return false
+	}
+	return true
+}
+
+// scrollIdl scrolls the screen n lines by using [ansi.DL] at del and using
+// [ansi.IL] at ins.
+func (s *Screen) scrollIdl(n, del, ins int, blank *Cell) bool {
+	if n < 0 {
+		return false
+	}
+
+	// Delete lines
+	s.move(0, del)
+	s.updatePen(blank)
+	s.buf.WriteString(ansi.DeleteLine(n))
+
+	// Insert lines
+	s.move(0, ins)
+	s.updatePen(blank)
+	s.buf.WriteString(ansi.InsertLine(n))
+
+	return true
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/hashmap.go b/vendor/github.com/charmbracelet/x/cellbuf/hashmap.go
new file mode 100644
index 0000000..0d25b54
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/hashmap.go
@@ -0,0 +1,301 @@
+package cellbuf
+
+import (
+	"github.com/charmbracelet/x/ansi"
+)
+
+// hash returns the hash value of a [Line].
+func hash(l Line) (h uint64) {
+	for _, c := range l {
+		var r rune
+		if c == nil {
+			r = ansi.SP
+		} else {
+			r = c.Rune
+		}
+		h += (h << 5) + uint64(r)
+	}
+	return
+}
+
+// hashmap represents a single [Line] hash.
+type hashmap struct {
+	value              uint64
+	oldcount, newcount int
+	oldindex, newindex int
+}
+
+// The value used to indicate lines created by insertions and scrolls.
+const newIndex = -1
+
+// updateHashmap updates the hashmap with the new hash value.
+func (s *Screen) updateHashmap() {
+	height := s.newbuf.Height()
+	if len(s.oldhash) >= height && len(s.newhash) >= height {
+		// rehash changed lines
+		for i := 0; i < height; i++ {
+			_, ok := s.touch[i]
+			if ok {
+				s.oldhash[i] = hash(s.curbuf.Line(i))
+				s.newhash[i] = hash(s.newbuf.Line(i))
+			}
+		}
+	} else {
+		// rehash all
+		if len(s.oldhash) != height {
+			s.oldhash = make([]uint64, height)
+		}
+		if len(s.newhash) != height {
+			s.newhash = make([]uint64, height)
+		}
+		for i := 0; i < height; i++ {
+			s.oldhash[i] = hash(s.curbuf.Line(i))
+			s.newhash[i] = hash(s.newbuf.Line(i))
+		}
+	}
+
+	s.hashtab = make([]hashmap, height*2)
+	for i := 0; i < height; i++ {
+		hashval := s.oldhash[i]
+
+		// Find matching hash or empty slot
+		idx := 0
+		for idx < len(s.hashtab) && s.hashtab[idx].value != 0 {
+			if s.hashtab[idx].value == hashval {
+				break
+			}
+			idx++
+		}
+
+		s.hashtab[idx].value = hashval // in case this is a new hash
+		s.hashtab[idx].oldcount++
+		s.hashtab[idx].oldindex = i
+	}
+	for i := 0; i < height; i++ {
+		hashval := s.newhash[i]
+
+		// Find matching hash or empty slot
+		idx := 0
+		for idx < len(s.hashtab) && s.hashtab[idx].value != 0 {
+			if s.hashtab[idx].value == hashval {
+				break
+			}
+			idx++
+		}
+
+		s.hashtab[idx].value = hashval // in case this is a new hash
+		s.hashtab[idx].newcount++
+		s.hashtab[idx].newindex = i
+
+		s.oldnum[i] = newIndex // init old indices slice
+	}
+
+	// Mark line pair corresponding to unique hash pairs.
+	for i := 0; i < len(s.hashtab) && s.hashtab[i].value != 0; i++ {
+		hsp := &s.hashtab[i]
+		if hsp.oldcount == 1 && hsp.newcount == 1 && hsp.oldindex != hsp.newindex {
+			s.oldnum[hsp.newindex] = hsp.oldindex
+		}
+	}
+
+	s.growHunks()
+
+	// Eliminate bad or impossible shifts. This includes removing those hunks
+	// which could not grow because of conflicts, as well those which are to be
+	// moved too far, they are likely to destroy more than carry.
+	for i := 0; i < height; {
+		var start, shift, size int
+		for i < height && s.oldnum[i] == newIndex {
+			i++
+		}
+		if i >= height {
+			break
+		}
+		start = i
+		shift = s.oldnum[i] - i
+		i++
+		for i < height && s.oldnum[i] != newIndex && s.oldnum[i]-i == shift {
+			i++
+		}
+		size = i - start
+		if size < 3 || size+min(size/8, 2) < abs(shift) {
+			for start < i {
+				s.oldnum[start] = newIndex
+				start++
+			}
+		}
+	}
+
+	// After clearing invalid hunks, try grow the rest.
+	s.growHunks()
+}
+
+// scrollOldhash
+func (s *Screen) scrollOldhash(n, top, bot int) {
+	if len(s.oldhash) == 0 {
+		return
+	}
+
+	size := bot - top + 1 - abs(n)
+	if n > 0 {
+		// Move existing hashes up
+		copy(s.oldhash[top:], s.oldhash[top+n:top+n+size])
+		// Recalculate hashes for newly shifted-in lines
+		for i := bot; i > bot-n; i-- {
+			s.oldhash[i] = hash(s.curbuf.Line(i))
+		}
+	} else {
+		// Move existing hashes down
+		copy(s.oldhash[top-n:], s.oldhash[top:top+size])
+		// Recalculate hashes for newly shifted-in lines
+		for i := top; i < top-n; i++ {
+			s.oldhash[i] = hash(s.curbuf.Line(i))
+		}
+	}
+}
+
+func (s *Screen) growHunks() {
+	var (
+		backLimit    int // limits for cells to fill
+		backRefLimit int // limit for references
+		i            int
+		nextHunk     int
+	)
+
+	height := s.newbuf.Height()
+	for i < height && s.oldnum[i] == newIndex {
+		i++
+	}
+	for ; i < height; i = nextHunk {
+		var (
+			forwardLimit    int
+			forwardRefLimit int
+			end             int
+			start           = i
+			shift           = s.oldnum[i] - i
+		)
+
+		// get forward limit
+		i = start + 1
+		for i < height &&
+			s.oldnum[i] != newIndex &&
+			s.oldnum[i]-i == shift {
+			i++
+		}
+
+		end = i
+		for i < height && s.oldnum[i] == newIndex {
+			i++
+		}
+
+		nextHunk = i
+		forwardLimit = i
+		if i >= height || s.oldnum[i] >= i {
+			forwardRefLimit = i
+		} else {
+			forwardRefLimit = s.oldnum[i]
+		}
+
+		i = start - 1
+
+		// grow back
+		if shift < 0 {
+			backLimit = backRefLimit + (-shift)
+		}
+		for i >= backLimit {
+			if s.newhash[i] == s.oldhash[i+shift] ||
+				s.costEffective(i+shift, i, shift < 0) {
+				s.oldnum[i] = i + shift
+			} else {
+				break
+			}
+			i--
+		}
+
+		i = end
+		// grow forward
+		if shift > 0 {
+			forwardLimit = forwardRefLimit - shift
+		}
+		for i < forwardLimit {
+			if s.newhash[i] == s.oldhash[i+shift] ||
+				s.costEffective(i+shift, i, shift > 0) {
+				s.oldnum[i] = i + shift
+			} else {
+				break
+			}
+			i++
+		}
+
+		backLimit = i
+		backRefLimit = backLimit
+		if shift > 0 {
+			backRefLimit += shift
+		}
+	}
+}
+
+// costEffective returns true if the cost of moving line 'from' to line 'to' seems to be
+// cost effective. 'blank' indicates whether the line 'to' would become blank.
+func (s *Screen) costEffective(from, to int, blank bool) bool {
+	if from == to {
+		return false
+	}
+
+	newFrom := s.oldnum[from]
+	if newFrom == newIndex {
+		newFrom = from
+	}
+
+	// On the left side of >= is the cost before moving. On the right side --
+	// cost after moving.
+
+	// Calculate costs before moving.
+	var costBeforeMove int
+	if blank {
+		// Cost of updating blank line at destination.
+		costBeforeMove = s.updateCostBlank(s.newbuf.Line(to))
+	} else {
+		// Cost of updating exiting line at destination.
+		costBeforeMove = s.updateCost(s.curbuf.Line(to), s.newbuf.Line(to))
+	}
+
+	// Add cost of updating source line
+	costBeforeMove += s.updateCost(s.curbuf.Line(newFrom), s.newbuf.Line(from))
+
+	// Calculate costs after moving.
+	var costAfterMove int
+	if newFrom == from {
+		// Source becomes blank after move
+		costAfterMove = s.updateCostBlank(s.newbuf.Line(from))
+	} else {
+		// Source gets updated from another line
+		costAfterMove = s.updateCost(s.curbuf.Line(newFrom), s.newbuf.Line(from))
+	}
+
+	// Add cost of moving source line to destination
+	costAfterMove += s.updateCost(s.curbuf.Line(from), s.newbuf.Line(to))
+
+	// Return true if moving is cost effective (costs less or equal)
+	return costBeforeMove >= costAfterMove
+}
+
+func (s *Screen) updateCost(from, to Line) (cost int) {
+	var fidx, tidx int
+	for i := s.newbuf.Width() - 1; i > 0; i, fidx, tidx = i-1, fidx+1, tidx+1 {
+		if !cellEqual(from.At(fidx), to.At(tidx)) {
+			cost++
+		}
+	}
+	return
+}
+
+func (s *Screen) updateCostBlank(to Line) (cost int) {
+	var tidx int
+	for i := s.newbuf.Width() - 1; i > 0; i, tidx = i-1, tidx+1 {
+		if !cellEqual(nil, to.At(tidx)) {
+			cost++
+		}
+	}
+	return
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/link.go b/vendor/github.com/charmbracelet/x/cellbuf/link.go
new file mode 100644
index 0000000..112f8e8
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/link.go
@@ -0,0 +1,14 @@
+package cellbuf
+
+import (
+	"github.com/charmbracelet/colorprofile"
+)
+
+// Convert converts a hyperlink to respect the given color profile.
+func ConvertLink(h Link, p colorprofile.Profile) Link {
+	if p == colorprofile.NoTTY {
+		return Link{}
+	}
+
+	return h
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/screen.go b/vendor/github.com/charmbracelet/x/cellbuf/screen.go
new file mode 100644
index 0000000..963b9ca
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/screen.go
@@ -0,0 +1,1457 @@
+package cellbuf
+
+import (
+	"bytes"
+	"errors"
+	"io"
+	"os"
+	"strings"
+	"sync"
+
+	"github.com/charmbracelet/colorprofile"
+	"github.com/charmbracelet/x/ansi"
+	"github.com/charmbracelet/x/term"
+)
+
+// ErrInvalidDimensions is returned when the dimensions of a window are invalid
+// for the operation.
+var ErrInvalidDimensions = errors.New("invalid dimensions")
+
+// notLocal returns whether the coordinates are not considered local movement
+// using the defined thresholds.
+// This takes the number of columns, and the coordinates of the current and
+// target positions.
+func notLocal(cols, fx, fy, tx, ty int) bool {
+	// The typical distance for a [ansi.CUP] sequence. Anything less than this
+	// is considered local movement.
+	const longDist = 8 - 1
+	return (tx > longDist) &&
+		(tx < cols-1-longDist) &&
+		(abs(ty-fy)+abs(tx-fx) > longDist)
+}
+
+// relativeCursorMove returns the relative cursor movement sequence using one or two
+// of the following sequences [ansi.CUU], [ansi.CUD], [ansi.CUF], [ansi.CUB],
+// [ansi.VPA], [ansi.HPA].
+// When overwrite is true, this will try to optimize the sequence by using the
+// screen cells values to move the cursor instead of using escape sequences.
+func relativeCursorMove(s *Screen, fx, fy, tx, ty int, overwrite, useTabs, useBackspace bool) string {
+	var seq strings.Builder
+
+	width, height := s.newbuf.Width(), s.newbuf.Height()
+	if ty != fy {
+		var yseq string
+		if s.xtermLike && !s.opts.RelativeCursor {
+			yseq = ansi.VerticalPositionAbsolute(ty + 1)
+		}
+
+		// OPTIM: Use [ansi.LF] and [ansi.ReverseIndex] as optimizations.
+
+		if ty > fy {
+			n := ty - fy
+			if cud := ansi.CursorDown(n); yseq == "" || len(cud) < len(yseq) {
+				yseq = cud
+			}
+			shouldScroll := !s.opts.AltScreen && fy+n >= s.scrollHeight
+			if lf := strings.Repeat("\n", n); shouldScroll || (fy+n < height && len(lf) < len(yseq)) {
+				// TODO: Ensure we're not unintentionally scrolling the screen down.
+				yseq = lf
+				s.scrollHeight = max(s.scrollHeight, fy+n)
+			}
+		} else if ty < fy {
+			n := fy - ty
+			if cuu := ansi.CursorUp(n); yseq == "" || len(cuu) < len(yseq) {
+				yseq = cuu
+			}
+			if n == 1 && fy-1 > 0 {
+				// TODO: Ensure we're not unintentionally scrolling the screen up.
+				yseq = ansi.ReverseIndex
+			}
+		}
+
+		seq.WriteString(yseq)
+	}
+
+	if tx != fx {
+		var xseq string
+		if s.xtermLike && !s.opts.RelativeCursor {
+			xseq = ansi.HorizontalPositionAbsolute(tx + 1)
+		}
+
+		if tx > fx {
+			n := tx - fx
+			if useTabs {
+				var tabs int
+				var col int
+				for col = fx; s.tabs.Next(col) <= tx; col = s.tabs.Next(col) {
+					tabs++
+					if col == s.tabs.Next(col) || col >= width-1 {
+						break
+					}
+				}
+
+				if tabs > 0 {
+					cht := ansi.CursorHorizontalForwardTab(tabs)
+					tab := strings.Repeat("\t", tabs)
+					if false && s.xtermLike && len(cht) < len(tab) {
+						// TODO: The linux console and some terminals such as
+						// Alacritty don't support [ansi.CHT]. Enable this when
+						// we have a way to detect this, or after 5 years when
+						// we're sure everyone has updated their terminals :P
+						seq.WriteString(cht)
+					} else {
+						seq.WriteString(tab)
+					}
+
+					n = tx - col
+					fx = col
+				}
+			}
+
+			if cuf := ansi.CursorForward(n); xseq == "" || len(cuf) < len(xseq) {
+				xseq = cuf
+			}
+
+			// If we have no attribute and style changes, overwrite is cheaper.
+			var ovw string
+			if overwrite && ty >= 0 {
+				for i := 0; i < n; i++ {
+					cell := s.newbuf.Cell(fx+i, ty)
+					if cell != nil && cell.Width > 0 {
+						i += cell.Width - 1
+						if !cell.Style.Equal(&s.cur.Style) || !cell.Link.Equal(&s.cur.Link) {
+							overwrite = false
+							break
+						}
+					}
+				}
+			}
+
+			if overwrite && ty >= 0 {
+				for i := 0; i < n; i++ {
+					cell := s.newbuf.Cell(fx+i, ty)
+					if cell != nil && cell.Width > 0 {
+						ovw += cell.String()
+						i += cell.Width - 1
+					} else {
+						ovw += " "
+					}
+				}
+			}
+
+			if overwrite && len(ovw) < len(xseq) {
+				xseq = ovw
+			}
+		} else if tx < fx {
+			n := fx - tx
+			if useTabs && s.xtermLike {
+				// VT100 does not support backward tabs [ansi.CBT].
+
+				col := fx
+
+				var cbt int // cursor backward tabs count
+				for s.tabs.Prev(col) >= tx {
+					col = s.tabs.Prev(col)
+					cbt++
+					if col == s.tabs.Prev(col) || col <= 0 {
+						break
+					}
+				}
+
+				if cbt > 0 {
+					seq.WriteString(ansi.CursorBackwardTab(cbt))
+					n = col - tx
+				}
+			}
+
+			if cub := ansi.CursorBackward(n); xseq == "" || len(cub) < len(xseq) {
+				xseq = cub
+			}
+
+			if useBackspace && n < len(xseq) {
+				xseq = strings.Repeat("\b", n)
+			}
+		}
+
+		seq.WriteString(xseq)
+	}
+
+	return seq.String()
+}
+
+// moveCursor moves and returns the cursor movement sequence to move the cursor
+// to the specified position.
+// When overwrite is true, this will try to optimize the sequence by using the
+// screen cells values to move the cursor instead of using escape sequences.
+func moveCursor(s *Screen, x, y int, overwrite bool) (seq string) {
+	fx, fy := s.cur.X, s.cur.Y
+
+	if !s.opts.RelativeCursor {
+		// Method #0: Use [ansi.CUP] if the distance is long.
+		seq = ansi.CursorPosition(x+1, y+1)
+		if fx == -1 || fy == -1 || notLocal(s.newbuf.Width(), fx, fy, x, y) {
+			return
+		}
+	}
+
+	// Optimize based on options.
+	trials := 0
+	if s.opts.HardTabs {
+		trials |= 2 // 0b10 in binary
+	}
+	if s.opts.Backspace {
+		trials |= 1 // 0b01 in binary
+	}
+
+	// Try all possible combinations of hard tabs and backspace optimizations.
+	for i := 0; i <= trials; i++ {
+		// Skip combinations that are not enabled.
+		if i & ^trials != 0 {
+			continue
+		}
+
+		useHardTabs := i&2 != 0
+		useBackspace := i&1 != 0
+
+		// Method #1: Use local movement sequences.
+		nseq := relativeCursorMove(s, fx, fy, x, y, overwrite, useHardTabs, useBackspace)
+		if (i == 0 && len(seq) == 0) || len(nseq) < len(seq) {
+			seq = nseq
+		}
+
+		// Method #2: Use [ansi.CR] and local movement sequences.
+		nseq = "\r" + relativeCursorMove(s, 0, fy, x, y, overwrite, useHardTabs, useBackspace)
+		if len(nseq) < len(seq) {
+			seq = nseq
+		}
+
+		if !s.opts.RelativeCursor {
+			// Method #3: Use [ansi.CursorHomePosition] and local movement sequences.
+			nseq = ansi.CursorHomePosition + relativeCursorMove(s, 0, 0, x, y, overwrite, useHardTabs, useBackspace)
+			if len(nseq) < len(seq) {
+				seq = nseq
+			}
+		}
+	}
+
+	return
+}
+
+// moveCursor moves the cursor to the specified position.
+func (s *Screen) moveCursor(x, y int, overwrite bool) {
+	if !s.opts.AltScreen && s.cur.X == -1 && s.cur.Y == -1 {
+		// First cursor movement in inline mode, move the cursor to the first
+		// column before moving to the target position.
+		s.buf.WriteByte('\r') //nolint:errcheck
+		s.cur.X, s.cur.Y = 0, 0
+	}
+	s.buf.WriteString(moveCursor(s, x, y, overwrite)) //nolint:errcheck
+	s.cur.X, s.cur.Y = x, y
+}
+
+func (s *Screen) move(x, y int) {
+	// XXX: Make sure we use the max height and width of the buffer in case
+	// we're in the middle of a resize operation.
+	width := max(s.newbuf.Width(), s.curbuf.Width())
+	height := max(s.newbuf.Height(), s.curbuf.Height())
+
+	if width > 0 && x >= width {
+		// Handle autowrap
+		y += (x / width)
+		x %= width
+	}
+
+	// XXX: Disable styles if there's any
+	// Some move operations such as [ansi.LF] can apply styles to the new
+	// cursor position, thus, we need to reset the styles before moving the
+	// cursor.
+	blank := s.clearBlank()
+	resetPen := y != s.cur.Y && !blank.Equal(&BlankCell)
+	if resetPen {
+		s.updatePen(nil)
+	}
+
+	// Reset wrap around (phantom cursor) state
+	if s.atPhantom {
+		s.cur.X = 0
+		s.buf.WriteByte('\r') //nolint:errcheck
+		s.atPhantom = false   // reset phantom cell state
+	}
+
+	// TODO: Investigate if we need to handle this case and/or if we need the
+	// following code.
+	//
+	// if width > 0 && s.cur.X >= width {
+	// 	l := (s.cur.X + 1) / width
+	//
+	// 	s.cur.Y += l
+	// 	if height > 0 && s.cur.Y >= height {
+	// 		l -= s.cur.Y - height - 1
+	// 	}
+	//
+	// 	if l > 0 {
+	// 		s.cur.X = 0
+	// 		s.buf.WriteString("\r" + strings.Repeat("\n", l)) //nolint:errcheck
+	// 	}
+	// }
+
+	if height > 0 {
+		if s.cur.Y > height-1 {
+			s.cur.Y = height - 1
+		}
+		if y > height-1 {
+			y = height - 1
+		}
+	}
+
+	if x == s.cur.X && y == s.cur.Y {
+		// We give up later because we need to run checks for the phantom cell
+		// and others before we can determine if we can give up.
+		return
+	}
+
+	// We set the new cursor in [Screen.moveCursor].
+	s.moveCursor(x, y, true) // Overwrite cells if possible
+}
+
+// Cursor represents a terminal Cursor.
+type Cursor struct {
+	Style
+	Link
+	Position
+}
+
+// ScreenOptions are options for the screen.
+type ScreenOptions struct {
+	// Term is the terminal type to use when writing to the screen. When empty,
+	// `$TERM` is used from [os.Getenv].
+	Term string
+	// Profile is the color profile to use when writing to the screen.
+	Profile colorprofile.Profile
+	// RelativeCursor is whether to use relative cursor movements. This is
+	// useful when alt-screen is not used or when using inline mode.
+	RelativeCursor bool
+	// AltScreen is whether to use the alternate screen buffer.
+	AltScreen bool
+	// ShowCursor is whether to show the cursor.
+	ShowCursor bool
+	// HardTabs is whether to use hard tabs to optimize cursor movements.
+	HardTabs bool
+	// Backspace is whether to use backspace characters to move the cursor.
+	Backspace bool
+}
+
+// lineData represents the metadata for a line.
+type lineData struct {
+	// first and last changed cell indices
+	firstCell, lastCell int
+	// old index used for scrolling
+	oldIndex int //nolint:unused
+}
+
+// Screen represents the terminal screen.
+type Screen struct {
+	w                io.Writer
+	buf              *bytes.Buffer // buffer for writing to the screen
+	curbuf           *Buffer       // the current buffer
+	newbuf           *Buffer       // the new buffer
+	tabs             *TabStops
+	touch            map[int]lineData
+	queueAbove       []string  // the queue of strings to write above the screen
+	oldhash, newhash []uint64  // the old and new hash values for each line
+	hashtab          []hashmap // the hashmap table
+	oldnum           []int     // old indices from previous hash
+	cur, saved       Cursor    // the current and saved cursors
+	opts             ScreenOptions
+	mu               sync.Mutex
+	method           ansi.Method
+	scrollHeight     int  // keeps track of how many lines we've scrolled down (inline mode)
+	altScreenMode    bool // whether alternate screen mode is enabled
+	cursorHidden     bool // whether text cursor mode is enabled
+	clear            bool // whether to force clear the screen
+	xtermLike        bool // whether to use xterm-like optimizations, otherwise, it uses vt100 only
+	queuedText       bool // whether we have queued non-zero width text queued up
+	atPhantom        bool // whether the cursor is out of bounds and at a phantom cell
+}
+
+// SetMethod sets the method used to calculate the width of cells.
+func (s *Screen) SetMethod(method ansi.Method) {
+	s.method = method
+}
+
+// UseBackspaces sets whether to use backspace characters to move the cursor.
+func (s *Screen) UseBackspaces(v bool) {
+	s.opts.Backspace = v
+}
+
+// UseHardTabs sets whether to use hard tabs to optimize cursor movements.
+func (s *Screen) UseHardTabs(v bool) {
+	s.opts.HardTabs = v
+}
+
+// SetColorProfile sets the color profile to use when writing to the screen.
+func (s *Screen) SetColorProfile(p colorprofile.Profile) {
+	s.opts.Profile = p
+}
+
+// SetRelativeCursor sets whether to use relative cursor movements.
+func (s *Screen) SetRelativeCursor(v bool) {
+	s.opts.RelativeCursor = v
+}
+
+// EnterAltScreen enters the alternate screen buffer.
+func (s *Screen) EnterAltScreen() {
+	s.opts.AltScreen = true
+	s.clear = true
+	s.saved = s.cur
+}
+
+// ExitAltScreen exits the alternate screen buffer.
+func (s *Screen) ExitAltScreen() {
+	s.opts.AltScreen = false
+	s.clear = true
+	s.cur = s.saved
+}
+
+// ShowCursor shows the cursor.
+func (s *Screen) ShowCursor() {
+	s.opts.ShowCursor = true
+}
+
+// HideCursor hides the cursor.
+func (s *Screen) HideCursor() {
+	s.opts.ShowCursor = false
+}
+
+// Bounds implements Window.
+func (s *Screen) Bounds() Rectangle {
+	// Always return the new buffer bounds.
+	return s.newbuf.Bounds()
+}
+
+// Cell implements Window.
+func (s *Screen) Cell(x int, y int) *Cell {
+	return s.newbuf.Cell(x, y)
+}
+
+// Redraw forces a full redraw of the screen.
+func (s *Screen) Redraw() {
+	s.mu.Lock()
+	s.clear = true
+	s.mu.Unlock()
+}
+
+// Clear clears the screen with blank cells. This is a convenience method for
+// [Screen.Fill] with a nil cell.
+func (s *Screen) Clear() bool {
+	return s.ClearRect(s.newbuf.Bounds())
+}
+
+// ClearRect clears the given rectangle with blank cells. This is a convenience
+// method for [Screen.FillRect] with a nil cell.
+func (s *Screen) ClearRect(r Rectangle) bool {
+	return s.FillRect(nil, r)
+}
+
+// SetCell implements Window.
+func (s *Screen) SetCell(x int, y int, cell *Cell) (v bool) {
+	s.mu.Lock()
+	defer s.mu.Unlock()
+	cellWidth := 1
+	if cell != nil {
+		cellWidth = cell.Width
+	}
+	if prev := s.curbuf.Cell(x, y); !cellEqual(prev, cell) {
+		chg, ok := s.touch[y]
+		if !ok {
+			chg = lineData{firstCell: x, lastCell: x + cellWidth}
+		} else {
+			chg.firstCell = min(chg.firstCell, x)
+			chg.lastCell = max(chg.lastCell, x+cellWidth)
+		}
+		s.touch[y] = chg
+	}
+
+	return s.newbuf.SetCell(x, y, cell)
+}
+
+// Fill implements Window.
+func (s *Screen) Fill(cell *Cell) bool {
+	return s.FillRect(cell, s.newbuf.Bounds())
+}
+
+// FillRect implements Window.
+func (s *Screen) FillRect(cell *Cell, r Rectangle) bool {
+	s.mu.Lock()
+	defer s.mu.Unlock()
+	s.newbuf.FillRect(cell, r)
+	for i := r.Min.Y; i < r.Max.Y; i++ {
+		s.touch[i] = lineData{firstCell: r.Min.X, lastCell: r.Max.X}
+	}
+	return true
+}
+
+// isXtermLike returns whether the terminal is xterm-like. This means that the
+// terminal supports ECMA-48 and ANSI X3.64 escape sequences.
+// TODO: Should this be a lookup table into each $TERM terminfo database? Like
+// we could keep a map of ANSI escape sequence to terminfo capability name and
+// check if the database supports the escape sequence. Instead of keeping a
+// list of terminal names here.
+func isXtermLike(termtype string) (v bool) {
+	parts := strings.Split(termtype, "-")
+	if len(parts) == 0 {
+		return
+	}
+
+	switch parts[0] {
+	case
+		"alacritty",
+		"contour",
+		"foot",
+		"ghostty",
+		"kitty",
+		"linux",
+		"rio",
+		"screen",
+		"st",
+		"tmux",
+		"wezterm",
+		"xterm":
+		v = true
+	}
+
+	return
+}
+
+// NewScreen creates a new Screen.
+func NewScreen(w io.Writer, width, height int, opts *ScreenOptions) (s *Screen) {
+	s = new(Screen)
+	s.w = w
+	if opts != nil {
+		s.opts = *opts
+	}
+
+	if s.opts.Term == "" {
+		s.opts.Term = os.Getenv("TERM")
+	}
+
+	if width <= 0 || height <= 0 {
+		if f, ok := w.(term.File); ok {
+			width, height, _ = term.GetSize(f.Fd())
+		}
+	}
+	if width < 0 {
+		width = 0
+	}
+	if height < 0 {
+		height = 0
+	}
+
+	s.buf = new(bytes.Buffer)
+	s.xtermLike = isXtermLike(s.opts.Term)
+	s.curbuf = NewBuffer(width, height)
+	s.newbuf = NewBuffer(width, height)
+	s.cur = Cursor{Position: Pos(-1, -1)} // start at -1 to force a move
+	s.saved = s.cur
+	s.reset()
+
+	return
+}
+
+// Width returns the width of the screen.
+func (s *Screen) Width() int {
+	return s.newbuf.Width()
+}
+
+// Height returns the height of the screen.
+func (s *Screen) Height() int {
+	return s.newbuf.Height()
+}
+
+// cellEqual returns whether the two cells are equal. A nil cell is considered
+// a [BlankCell].
+func cellEqual(a, b *Cell) bool {
+	if a == b {
+		return true
+	}
+	if a == nil {
+		a = &BlankCell
+	}
+	if b == nil {
+		b = &BlankCell
+	}
+	return a.Equal(b)
+}
+
+// putCell draws a cell at the current cursor position.
+func (s *Screen) putCell(cell *Cell) {
+	width, height := s.newbuf.Width(), s.newbuf.Height()
+	if s.opts.AltScreen && s.cur.X == width-1 && s.cur.Y == height-1 {
+		s.putCellLR(cell)
+	} else {
+		s.putAttrCell(cell)
+	}
+}
+
+// wrapCursor wraps the cursor to the next line.
+//
+//nolint:unused
+func (s *Screen) wrapCursor() {
+	const autoRightMargin = true
+	if autoRightMargin {
+		// Assume we have auto wrap mode enabled.
+		s.cur.X = 0
+		s.cur.Y++
+	} else {
+		s.cur.X--
+	}
+}
+
+func (s *Screen) putAttrCell(cell *Cell) {
+	if cell != nil && cell.Empty() {
+		// XXX: Zero width cells are special and should not be written to the
+		// screen no matter what other attributes they have.
+		// Zero width cells are used for wide characters that are split into
+		// multiple cells.
+		return
+	}
+
+	if cell == nil {
+		cell = s.clearBlank()
+	}
+
+	// We're at pending wrap state (phantom cell), incoming cell should
+	// wrap.
+	if s.atPhantom {
+		s.wrapCursor()
+		s.atPhantom = false
+	}
+
+	s.updatePen(cell)
+	s.buf.WriteRune(cell.Rune) //nolint:errcheck
+	for _, c := range cell.Comb {
+		s.buf.WriteRune(c) //nolint:errcheck
+	}
+
+	s.cur.X += cell.Width
+
+	if cell.Width > 0 {
+		s.queuedText = true
+	}
+
+	if s.cur.X >= s.newbuf.Width() {
+		s.atPhantom = true
+	}
+}
+
+// putCellLR draws a cell at the lower right corner of the screen.
+func (s *Screen) putCellLR(cell *Cell) {
+	// Optimize for the lower right corner cell.
+	curX := s.cur.X
+	if cell == nil || !cell.Empty() {
+		s.buf.WriteString(ansi.ResetAutoWrapMode) //nolint:errcheck
+		s.putAttrCell(cell)
+		// Writing to lower-right corner cell should not wrap.
+		s.atPhantom = false
+		s.cur.X = curX
+		s.buf.WriteString(ansi.SetAutoWrapMode) //nolint:errcheck
+	}
+}
+
+// updatePen updates the cursor pen styles.
+func (s *Screen) updatePen(cell *Cell) {
+	if cell == nil {
+		cell = &BlankCell
+	}
+
+	if s.opts.Profile != 0 {
+		// Downsample colors to the given color profile.
+		cell.Style = ConvertStyle(cell.Style, s.opts.Profile)
+		cell.Link = ConvertLink(cell.Link, s.opts.Profile)
+	}
+
+	if !cell.Style.Equal(&s.cur.Style) {
+		seq := cell.Style.DiffSequence(s.cur.Style)
+		if cell.Style.Empty() && len(seq) > len(ansi.ResetStyle) {
+			seq = ansi.ResetStyle
+		}
+		s.buf.WriteString(seq) //nolint:errcheck
+		s.cur.Style = cell.Style
+	}
+	if !cell.Link.Equal(&s.cur.Link) {
+		s.buf.WriteString(ansi.SetHyperlink(cell.Link.URL, cell.Link.Params)) //nolint:errcheck
+		s.cur.Link = cell.Link
+	}
+}
+
+// emitRange emits a range of cells to the buffer. It it equivalent to calling
+// [Screen.putCell] for each cell in the range. This is optimized to use
+// [ansi.ECH] and [ansi.REP].
+// Returns whether the cursor is at the end of interval or somewhere in the
+// middle.
+func (s *Screen) emitRange(line Line, n int) (eoi bool) {
+	for n > 0 {
+		var count int
+		for n > 1 && !cellEqual(line.At(0), line.At(1)) {
+			s.putCell(line.At(0))
+			line = line[1:]
+			n--
+		}
+
+		cell0 := line[0]
+		if n == 1 {
+			s.putCell(cell0)
+			return false
+		}
+
+		count = 2
+		for count < n && cellEqual(line.At(count), cell0) {
+			count++
+		}
+
+		ech := ansi.EraseCharacter(count)
+		cup := ansi.CursorPosition(s.cur.X+count, s.cur.Y)
+		rep := ansi.RepeatPreviousCharacter(count)
+		if s.xtermLike && count > len(ech)+len(cup) && cell0 != nil && cell0.Clear() {
+			s.updatePen(cell0)
+			s.buf.WriteString(ech) //nolint:errcheck
+
+			// If this is the last cell, we don't need to move the cursor.
+			if count < n {
+				s.move(s.cur.X+count, s.cur.Y)
+			} else {
+				return true // cursor in the middle
+			}
+		} else if s.xtermLike && count > len(rep) &&
+			(cell0 == nil || (len(cell0.Comb) == 0 && cell0.Rune < 256)) {
+			// We only support ASCII characters. Most terminals will handle
+			// non-ASCII characters correctly, but some might not, ahem xterm.
+			//
+			// NOTE: [ansi.REP] only repeats the last rune and won't work
+			// if the last cell contains multiple runes.
+
+			wrapPossible := s.cur.X+count >= s.newbuf.Width()
+			repCount := count
+			if wrapPossible {
+				repCount--
+			}
+
+			s.updatePen(cell0)
+			s.putCell(cell0)
+			repCount-- // cell0 is a single width cell ASCII character
+
+			s.buf.WriteString(ansi.RepeatPreviousCharacter(repCount)) //nolint:errcheck
+			s.cur.X += repCount
+			if wrapPossible {
+				s.putCell(cell0)
+			}
+		} else {
+			for i := 0; i < count; i++ {
+				s.putCell(line.At(i))
+			}
+		}
+
+		line = line[clamp(count, 0, len(line)):]
+		n -= count
+	}
+
+	return
+}
+
+// putRange puts a range of cells from the old line to the new line.
+// Returns whether the cursor is at the end of interval or somewhere in the
+// middle.
+func (s *Screen) putRange(oldLine, newLine Line, y, start, end int) (eoi bool) {
+	inline := min(len(ansi.CursorPosition(start+1, y+1)),
+		min(len(ansi.HorizontalPositionAbsolute(start+1)),
+			len(ansi.CursorForward(start+1))))
+	if (end - start + 1) > inline {
+		var j, same int
+		for j, same = start, 0; j <= end; j++ {
+			oldCell, newCell := oldLine.At(j), newLine.At(j)
+			if same == 0 && oldCell != nil && oldCell.Empty() {
+				continue
+			}
+			if cellEqual(oldCell, newCell) {
+				same++
+			} else {
+				if same > end-start {
+					s.emitRange(newLine[start:], j-same-start)
+					s.move(j, y)
+					start = j
+				}
+				same = 0
+			}
+		}
+
+		i := s.emitRange(newLine[start:], j-same-start)
+
+		// Always return 1 for the next [Screen.move] after a [Screen.putRange] if
+		// we found identical characters at end of interval.
+		if same == 0 {
+			return i
+		}
+		return true
+	}
+
+	return s.emitRange(newLine[start:], end-start+1)
+}
+
+// clearToEnd clears the screen from the current cursor position to the end of
+// line.
+func (s *Screen) clearToEnd(blank *Cell, force bool) { //nolint:unparam
+	if s.cur.Y >= 0 {
+		curline := s.curbuf.Line(s.cur.Y)
+		for j := s.cur.X; j < s.curbuf.Width(); j++ {
+			if j >= 0 {
+				c := curline.At(j)
+				if !cellEqual(c, blank) {
+					curline.Set(j, blank)
+					force = true
+				}
+			}
+		}
+	}
+
+	if force {
+		s.updatePen(blank)
+		count := s.newbuf.Width() - s.cur.X
+		if s.el0Cost() <= count {
+			s.buf.WriteString(ansi.EraseLineRight) //nolint:errcheck
+		} else {
+			for i := 0; i < count; i++ {
+				s.putCell(blank)
+			}
+		}
+	}
+}
+
+// clearBlank returns a blank cell based on the current cursor background color.
+func (s *Screen) clearBlank() *Cell {
+	c := BlankCell
+	if !s.cur.Style.Empty() || !s.cur.Link.Empty() {
+		c.Style = s.cur.Style
+		c.Link = s.cur.Link
+	}
+	return &c
+}
+
+// insertCells inserts the count cells pointed by the given line at the current
+// cursor position.
+func (s *Screen) insertCells(line Line, count int) {
+	if s.xtermLike {
+		// Use [ansi.ICH] as an optimization.
+		s.buf.WriteString(ansi.InsertCharacter(count)) //nolint:errcheck
+	} else {
+		// Otherwise, use [ansi.IRM] mode.
+		s.buf.WriteString(ansi.SetInsertReplaceMode) //nolint:errcheck
+	}
+
+	for i := 0; count > 0; i++ {
+		s.putAttrCell(line[i])
+		count--
+	}
+
+	if !s.xtermLike {
+		s.buf.WriteString(ansi.ResetInsertReplaceMode) //nolint:errcheck
+	}
+}
+
+// el0Cost returns the cost of using [ansi.EL] 0 i.e. [ansi.EraseLineRight]. If
+// this terminal supports background color erase, it can be cheaper to use
+// [ansi.EL] 0 i.e. [ansi.EraseLineRight] to clear
+// trailing spaces.
+func (s *Screen) el0Cost() int {
+	if s.xtermLike {
+		return 0
+	}
+	return len(ansi.EraseLineRight)
+}
+
+// transformLine transforms the given line in the current window to the
+// corresponding line in the new window. It uses [ansi.ICH] and [ansi.DCH] to
+// insert or delete characters.
+func (s *Screen) transformLine(y int) {
+	var firstCell, oLastCell, nLastCell int // first, old last, new last index
+	oldLine := s.curbuf.Line(y)
+	newLine := s.newbuf.Line(y)
+
+	// Find the first changed cell in the line
+	var lineChanged bool
+	for i := 0; i < s.newbuf.Width(); i++ {
+		if !cellEqual(newLine.At(i), oldLine.At(i)) {
+			lineChanged = true
+			break
+		}
+	}
+
+	const ceolStandoutGlitch = false
+	if ceolStandoutGlitch && lineChanged {
+		s.move(0, y)
+		s.clearToEnd(nil, false)
+		s.putRange(oldLine, newLine, y, 0, s.newbuf.Width()-1)
+	} else {
+		blank := newLine.At(0)
+
+		// It might be cheaper to clear leading spaces with [ansi.EL] 1 i.e.
+		// [ansi.EraseLineLeft].
+		if blank == nil || blank.Clear() {
+			var oFirstCell, nFirstCell int
+			for oFirstCell = 0; oFirstCell < s.curbuf.Width(); oFirstCell++ {
+				if !cellEqual(oldLine.At(oFirstCell), blank) {
+					break
+				}
+			}
+			for nFirstCell = 0; nFirstCell < s.newbuf.Width(); nFirstCell++ {
+				if !cellEqual(newLine.At(nFirstCell), blank) {
+					break
+				}
+			}
+
+			if nFirstCell == oFirstCell {
+				firstCell = nFirstCell
+
+				// Find the first differing cell
+				for firstCell < s.newbuf.Width() &&
+					cellEqual(oldLine.At(firstCell), newLine.At(firstCell)) {
+					firstCell++
+				}
+			} else if oFirstCell > nFirstCell {
+				firstCell = nFirstCell
+			} else if oFirstCell < nFirstCell {
+				firstCell = oFirstCell
+				el1Cost := len(ansi.EraseLineLeft)
+				if el1Cost < nFirstCell-oFirstCell {
+					if nFirstCell >= s.newbuf.Width() {
+						s.move(0, y)
+						s.updatePen(blank)
+						s.buf.WriteString(ansi.EraseLineRight) //nolint:errcheck
+					} else {
+						s.move(nFirstCell-1, y)
+						s.updatePen(blank)
+						s.buf.WriteString(ansi.EraseLineLeft) //nolint:errcheck
+					}
+
+					for firstCell < nFirstCell {
+						oldLine.Set(firstCell, blank)
+						firstCell++
+					}
+				}
+			}
+		} else {
+			// Find the first differing cell
+			for firstCell < s.newbuf.Width() && cellEqual(newLine.At(firstCell), oldLine.At(firstCell)) {
+				firstCell++
+			}
+		}
+
+		// If we didn't find one, we're done
+		if firstCell >= s.newbuf.Width() {
+			return
+		}
+
+		blank = newLine.At(s.newbuf.Width() - 1)
+		if blank != nil && !blank.Clear() {
+			// Find the last differing cell
+			nLastCell = s.newbuf.Width() - 1
+			for nLastCell > firstCell && cellEqual(newLine.At(nLastCell), oldLine.At(nLastCell)) {
+				nLastCell--
+			}
+
+			if nLastCell >= firstCell {
+				s.move(firstCell, y)
+				s.putRange(oldLine, newLine, y, firstCell, nLastCell)
+				if firstCell < len(oldLine) && firstCell < len(newLine) {
+					copy(oldLine[firstCell:], newLine[firstCell:])
+				} else {
+					copy(oldLine, newLine)
+				}
+			}
+
+			return
+		}
+
+		// Find last non-blank cell in the old line.
+		oLastCell = s.curbuf.Width() - 1
+		for oLastCell > firstCell && cellEqual(oldLine.At(oLastCell), blank) {
+			oLastCell--
+		}
+
+		// Find last non-blank cell in the new line.
+		nLastCell = s.newbuf.Width() - 1
+		for nLastCell > firstCell && cellEqual(newLine.At(nLastCell), blank) {
+			nLastCell--
+		}
+
+		if nLastCell == firstCell && s.el0Cost() < oLastCell-nLastCell {
+			s.move(firstCell, y)
+			if !cellEqual(newLine.At(firstCell), blank) {
+				s.putCell(newLine.At(firstCell))
+			}
+			s.clearToEnd(blank, false)
+		} else if nLastCell != oLastCell &&
+			!cellEqual(newLine.At(nLastCell), oldLine.At(oLastCell)) {
+			s.move(firstCell, y)
+			if oLastCell-nLastCell > s.el0Cost() {
+				if s.putRange(oldLine, newLine, y, firstCell, nLastCell) {
+					s.move(nLastCell+1, y)
+				}
+				s.clearToEnd(blank, false)
+			} else {
+				n := max(nLastCell, oLastCell)
+				s.putRange(oldLine, newLine, y, firstCell, n)
+			}
+		} else {
+			nLastNonBlank := nLastCell
+			oLastNonBlank := oLastCell
+
+			// Find the last cells that really differ.
+			// Can be -1 if no cells differ.
+			for cellEqual(newLine.At(nLastCell), oldLine.At(oLastCell)) {
+				if !cellEqual(newLine.At(nLastCell-1), oldLine.At(oLastCell-1)) {
+					break
+				}
+				nLastCell--
+				oLastCell--
+				if nLastCell == -1 || oLastCell == -1 {
+					break
+				}
+			}
+
+			n := min(oLastCell, nLastCell)
+			if n >= firstCell {
+				s.move(firstCell, y)
+				s.putRange(oldLine, newLine, y, firstCell, n)
+			}
+
+			if oLastCell < nLastCell {
+				m := max(nLastNonBlank, oLastNonBlank)
+				if n != 0 {
+					for n > 0 {
+						wide := newLine.At(n + 1)
+						if wide == nil || !wide.Empty() {
+							break
+						}
+						n--
+						oLastCell--
+					}
+				} else if n >= firstCell && newLine.At(n) != nil && newLine.At(n).Width > 1 {
+					next := newLine.At(n + 1)
+					for next != nil && next.Empty() {
+						n++
+						oLastCell++
+					}
+				}
+
+				s.move(n+1, y)
+				ichCost := 3 + nLastCell - oLastCell
+				if s.xtermLike && (nLastCell < nLastNonBlank || ichCost > (m-n)) {
+					s.putRange(oldLine, newLine, y, n+1, m)
+				} else {
+					s.insertCells(newLine[n+1:], nLastCell-oLastCell)
+				}
+			} else if oLastCell > nLastCell {
+				s.move(n+1, y)
+				dchCost := 3 + oLastCell - nLastCell
+				if dchCost > len(ansi.EraseLineRight)+nLastNonBlank-(n+1) {
+					if s.putRange(oldLine, newLine, y, n+1, nLastNonBlank) {
+						s.move(nLastNonBlank+1, y)
+					}
+					s.clearToEnd(blank, false)
+				} else {
+					s.updatePen(blank)
+					s.deleteCells(oLastCell - nLastCell)
+				}
+			}
+		}
+	}
+
+	// Update the old line with the new line
+	if firstCell < len(oldLine) && firstCell < len(newLine) {
+		copy(oldLine[firstCell:], newLine[firstCell:])
+	} else {
+		copy(oldLine, newLine)
+	}
+}
+
+// deleteCells deletes the count cells at the current cursor position and moves
+// the rest of the line to the left. This is equivalent to [ansi.DCH].
+func (s *Screen) deleteCells(count int) {
+	// [ansi.DCH] will shift in cells from the right margin so we need to
+	// ensure that they are the right style.
+	s.buf.WriteString(ansi.DeleteCharacter(count)) //nolint:errcheck
+}
+
+// clearToBottom clears the screen from the current cursor position to the end
+// of the screen.
+func (s *Screen) clearToBottom(blank *Cell) {
+	row, col := s.cur.Y, s.cur.X
+	if row < 0 {
+		row = 0
+	}
+
+	s.updatePen(blank)
+	s.buf.WriteString(ansi.EraseScreenBelow) //nolint:errcheck
+	// Clear the rest of the current line
+	s.curbuf.ClearRect(Rect(col, row, s.curbuf.Width()-col, 1))
+	// Clear everything below the current line
+	s.curbuf.ClearRect(Rect(0, row+1, s.curbuf.Width(), s.curbuf.Height()-row-1))
+}
+
+// clearBottom tests if clearing the end of the screen would satisfy part of
+// the screen update. Scan backwards through lines in the screen checking if
+// each is blank and one or more are changed.
+// It returns the top line.
+func (s *Screen) clearBottom(total int) (top int) {
+	if total <= 0 {
+		return
+	}
+
+	top = total
+	last := s.newbuf.Width()
+	blank := s.clearBlank()
+	canClearWithBlank := blank == nil || blank.Clear()
+
+	if canClearWithBlank {
+		var row int
+		for row = total - 1; row >= 0; row-- {
+			oldLine := s.curbuf.Line(row)
+			newLine := s.newbuf.Line(row)
+
+			var col int
+			ok := true
+			for col = 0; ok && col < last; col++ {
+				ok = cellEqual(newLine.At(col), blank)
+			}
+			if !ok {
+				break
+			}
+
+			for col = 0; ok && col < last; col++ {
+				ok = len(oldLine) == last && cellEqual(oldLine.At(col), blank)
+			}
+			if !ok {
+				top = row
+			}
+		}
+
+		if top < total {
+			s.move(0, top-1) // top is 1-based
+			s.clearToBottom(blank)
+			if s.oldhash != nil && s.newhash != nil &&
+				row < len(s.oldhash) && row < len(s.newhash) {
+				for row := top; row < s.newbuf.Height(); row++ {
+					s.oldhash[row] = s.newhash[row]
+				}
+			}
+		}
+	}
+
+	return
+}
+
+// clearScreen clears the screen and put cursor at home.
+func (s *Screen) clearScreen(blank *Cell) {
+	s.updatePen(blank)
+	s.buf.WriteString(ansi.CursorHomePosition) //nolint:errcheck
+	s.buf.WriteString(ansi.EraseEntireScreen)  //nolint:errcheck
+	s.cur.X, s.cur.Y = 0, 0
+	s.curbuf.Fill(blank)
+}
+
+// clearBelow clears everything below and including the row.
+func (s *Screen) clearBelow(blank *Cell, row int) {
+	s.move(0, row)
+	s.clearToBottom(blank)
+}
+
+// clearUpdate forces a screen redraw.
+func (s *Screen) clearUpdate() {
+	blank := s.clearBlank()
+	var nonEmpty int
+	if s.opts.AltScreen {
+		// XXX: We're using the maximum height of the two buffers to ensure
+		// we write newly added lines to the screen in [Screen.transformLine].
+		nonEmpty = max(s.curbuf.Height(), s.newbuf.Height())
+		s.clearScreen(blank)
+	} else {
+		nonEmpty = s.newbuf.Height()
+		s.clearBelow(blank, 0)
+	}
+	nonEmpty = s.clearBottom(nonEmpty)
+	for i := 0; i < nonEmpty; i++ {
+		s.transformLine(i)
+	}
+}
+
+// Flush flushes the buffer to the screen.
+func (s *Screen) Flush() (err error) {
+	s.mu.Lock()
+	defer s.mu.Unlock()
+	return s.flush()
+}
+
+func (s *Screen) flush() (err error) {
+	// Write the buffer
+	if s.buf.Len() > 0 {
+		_, err = s.w.Write(s.buf.Bytes()) //nolint:errcheck
+		if err == nil {
+			s.buf.Reset()
+		}
+	}
+
+	return
+}
+
+// Render renders changes of the screen to the internal buffer. Call
+// [Screen.Flush] to flush pending changes to the screen.
+func (s *Screen) Render() {
+	s.mu.Lock()
+	s.render()
+	s.mu.Unlock()
+}
+
+func (s *Screen) render() {
+	// Do we need to render anything?
+	if s.opts.AltScreen == s.altScreenMode &&
+		!s.opts.ShowCursor == s.cursorHidden &&
+		!s.clear &&
+		len(s.touch) == 0 &&
+		len(s.queueAbove) == 0 {
+		return
+	}
+
+	// TODO: Investigate whether this is necessary. Theoretically, terminals
+	// can add/remove tab stops and we should be able to handle that. We could
+	// use [ansi.DECTABSR] to read the tab stops, but that's not implemented in
+	// most terminals :/
+	// // Are we using hard tabs? If so, ensure tabs are using the
+	// // default interval using [ansi.DECST8C].
+	// if s.opts.HardTabs && !s.initTabs {
+	// 	s.buf.WriteString(ansi.SetTabEvery8Columns)
+	// 	s.initTabs = true
+	// }
+
+	// Do we need alt-screen mode?
+	if s.opts.AltScreen != s.altScreenMode {
+		if s.opts.AltScreen {
+			s.buf.WriteString(ansi.SetAltScreenSaveCursorMode)
+		} else {
+			s.buf.WriteString(ansi.ResetAltScreenSaveCursorMode)
+		}
+		s.altScreenMode = s.opts.AltScreen
+	}
+
+	// Do we need text cursor mode?
+	if !s.opts.ShowCursor != s.cursorHidden {
+		s.cursorHidden = !s.opts.ShowCursor
+		if s.cursorHidden {
+			s.buf.WriteString(ansi.HideCursor)
+		}
+	}
+
+	// Do we have queued strings to write above the screen?
+	if len(s.queueAbove) > 0 {
+		// TODO: Use scrolling region if available.
+		// TODO: Use [Screen.Write] [io.Writer] interface.
+
+		// We need to scroll the screen up by the number of lines in the queue.
+		// We can't use [ansi.SU] because we want the cursor to move down until
+		// it reaches the bottom of the screen.
+		s.move(0, s.newbuf.Height()-1)
+		s.buf.WriteString(strings.Repeat("\n", len(s.queueAbove)))
+		s.cur.Y += len(s.queueAbove)
+		// XXX: Now go to the top of the screen, insert new lines, and write
+		// the queued strings. It is important to use [Screen.moveCursor]
+		// instead of [Screen.move] because we don't want to perform any checks
+		// on the cursor position.
+		s.moveCursor(0, 0, false)
+		s.buf.WriteString(ansi.InsertLine(len(s.queueAbove)))
+		for _, line := range s.queueAbove {
+			s.buf.WriteString(line + "\r\n")
+		}
+
+		// Clear the queue
+		s.queueAbove = s.queueAbove[:0]
+	}
+
+	var nonEmpty int
+
+	// XXX: In inline mode, after a screen resize, we need to clear the extra
+	// lines at the bottom of the screen. This is because in inline mode, we
+	// don't use the full screen height and the current buffer size might be
+	// larger than the new buffer size.
+	partialClear := !s.opts.AltScreen && s.cur.X != -1 && s.cur.Y != -1 &&
+		s.curbuf.Width() == s.newbuf.Width() &&
+		s.curbuf.Height() > 0 &&
+		s.curbuf.Height() > s.newbuf.Height()
+
+	if !s.clear && partialClear {
+		s.clearBelow(nil, s.newbuf.Height()-1)
+	}
+
+	if s.clear {
+		s.clearUpdate()
+		s.clear = false
+	} else if len(s.touch) > 0 {
+		if s.opts.AltScreen {
+			// Optimize scrolling for the alternate screen buffer.
+			// TODO: Should we optimize for inline mode as well? If so, we need
+			// to know the actual cursor position to use [ansi.DECSTBM].
+			s.scrollOptimize()
+		}
+
+		var changedLines int
+		var i int
+
+		if s.opts.AltScreen {
+			nonEmpty = min(s.curbuf.Height(), s.newbuf.Height())
+		} else {
+			nonEmpty = s.newbuf.Height()
+		}
+
+		nonEmpty = s.clearBottom(nonEmpty)
+		for i = 0; i < nonEmpty; i++ {
+			_, ok := s.touch[i]
+			if ok {
+				s.transformLine(i)
+				changedLines++
+			}
+		}
+	}
+
+	// Sync windows and screen
+	s.touch = make(map[int]lineData, s.newbuf.Height())
+
+	if s.curbuf.Width() != s.newbuf.Width() || s.curbuf.Height() != s.newbuf.Height() {
+		// Resize the old buffer to match the new buffer.
+		_, oldh := s.curbuf.Width(), s.curbuf.Height()
+		s.curbuf.Resize(s.newbuf.Width(), s.newbuf.Height())
+		// Sync new lines to old lines
+		for i := oldh - 1; i < s.newbuf.Height(); i++ {
+			copy(s.curbuf.Line(i), s.newbuf.Line(i))
+		}
+	}
+
+	s.updatePen(nil) // nil indicates a blank cell with no styles
+
+	// Do we have enough changes to justify toggling the cursor?
+	if s.buf.Len() > 1 && s.opts.ShowCursor && !s.cursorHidden && s.queuedText {
+		nb := new(bytes.Buffer)
+		nb.Grow(s.buf.Len() + len(ansi.HideCursor) + len(ansi.ShowCursor))
+		nb.WriteString(ansi.HideCursor)
+		nb.Write(s.buf.Bytes())
+		nb.WriteString(ansi.ShowCursor)
+		*s.buf = *nb
+	}
+
+	s.queuedText = false
+}
+
+// Close writes the final screen update and resets the screen.
+func (s *Screen) Close() (err error) {
+	s.mu.Lock()
+	defer s.mu.Unlock()
+
+	s.render()
+	s.updatePen(nil)
+	// Go to the bottom of the screen
+	s.move(0, s.newbuf.Height()-1)
+
+	if s.altScreenMode {
+		s.buf.WriteString(ansi.ResetAltScreenSaveCursorMode)
+		s.altScreenMode = false
+	}
+
+	if s.cursorHidden {
+		s.buf.WriteString(ansi.ShowCursor)
+		s.cursorHidden = false
+	}
+
+	// Write the buffer
+	err = s.flush()
+	if err != nil {
+		return
+	}
+
+	s.reset()
+	return
+}
+
+// reset resets the screen to its initial state.
+func (s *Screen) reset() {
+	s.scrollHeight = 0
+	s.cursorHidden = false
+	s.altScreenMode = false
+	s.touch = make(map[int]lineData, s.newbuf.Height())
+	if s.curbuf != nil {
+		s.curbuf.Clear()
+	}
+	if s.newbuf != nil {
+		s.newbuf.Clear()
+	}
+	s.buf.Reset()
+	s.tabs = DefaultTabStops(s.newbuf.Width())
+	s.oldhash, s.newhash = nil, nil
+
+	// We always disable HardTabs when termtype is "linux".
+	if strings.HasPrefix(s.opts.Term, "linux") {
+		s.opts.HardTabs = false
+	}
+}
+
+// Resize resizes the screen.
+func (s *Screen) Resize(width, height int) bool {
+	oldw := s.newbuf.Width()
+	oldh := s.newbuf.Height()
+
+	if s.opts.AltScreen || width != oldw {
+		// We only clear the whole screen if the width changes. Adding/removing
+		// rows is handled by the [Screen.render] and [Screen.transformLine]
+		// methods.
+		s.clear = true
+	}
+
+	// Clear new columns and lines
+	if width > oldh {
+		s.ClearRect(Rect(max(oldw-1, 0), 0, width-oldw, height))
+	} else if width < oldw {
+		s.ClearRect(Rect(max(width-1, 0), 0, oldw-width, height))
+	}
+
+	if height > oldh {
+		s.ClearRect(Rect(0, max(oldh-1, 0), width, height-oldh))
+	} else if height < oldh {
+		s.ClearRect(Rect(0, max(height-1, 0), width, oldh-height))
+	}
+
+	s.mu.Lock()
+	s.newbuf.Resize(width, height)
+	s.tabs.Resize(width)
+	s.oldhash, s.newhash = nil, nil
+	s.scrollHeight = 0 // reset scroll lines
+	s.mu.Unlock()
+
+	return true
+}
+
+// MoveTo moves the cursor to the given position.
+func (s *Screen) MoveTo(x, y int) {
+	s.mu.Lock()
+	s.move(x, y)
+	s.mu.Unlock()
+}
+
+// InsertAbove inserts string above the screen. The inserted string is not
+// managed by the screen. This does nothing when alternate screen mode is
+// enabled.
+func (s *Screen) InsertAbove(str string) {
+	if s.opts.AltScreen {
+		return
+	}
+	s.mu.Lock()
+	for _, line := range strings.Split(str, "\n") {
+		s.queueAbove = append(s.queueAbove, s.method.Truncate(line, s.Width(), ""))
+	}
+	s.mu.Unlock()
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/sequence.go b/vendor/github.com/charmbracelet/x/cellbuf/sequence.go
new file mode 100644
index 0000000..613eefe
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/sequence.go
@@ -0,0 +1,131 @@
+package cellbuf
+
+import (
+	"bytes"
+	"image/color"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// ReadStyle reads a Select Graphic Rendition (SGR) escape sequences from a
+// list of parameters.
+func ReadStyle(params ansi.Params, pen *Style) {
+	if len(params) == 0 {
+		pen.Reset()
+		return
+	}
+
+	for i := 0; i < len(params); i++ {
+		param, hasMore, _ := params.Param(i, 0)
+		switch param {
+		case 0: // Reset
+			pen.Reset()
+		case 1: // Bold
+			pen.Bold(true)
+		case 2: // Dim/Faint
+			pen.Faint(true)
+		case 3: // Italic
+			pen.Italic(true)
+		case 4: // Underline
+			nextParam, _, ok := params.Param(i+1, 0)
+			if hasMore && ok { // Only accept subparameters i.e. separated by ":"
+				switch nextParam {
+				case 0, 1, 2, 3, 4, 5:
+					i++
+					switch nextParam {
+					case 0: // No Underline
+						pen.UnderlineStyle(NoUnderline)
+					case 1: // Single Underline
+						pen.UnderlineStyle(SingleUnderline)
+					case 2: // Double Underline
+						pen.UnderlineStyle(DoubleUnderline)
+					case 3: // Curly Underline
+						pen.UnderlineStyle(CurlyUnderline)
+					case 4: // Dotted Underline
+						pen.UnderlineStyle(DottedUnderline)
+					case 5: // Dashed Underline
+						pen.UnderlineStyle(DashedUnderline)
+					}
+				}
+			} else {
+				// Single Underline
+				pen.Underline(true)
+			}
+		case 5: // Slow Blink
+			pen.SlowBlink(true)
+		case 6: // Rapid Blink
+			pen.RapidBlink(true)
+		case 7: // Reverse
+			pen.Reverse(true)
+		case 8: // Conceal
+			pen.Conceal(true)
+		case 9: // Crossed-out/Strikethrough
+			pen.Strikethrough(true)
+		case 22: // Normal Intensity (not bold or faint)
+			pen.Bold(false).Faint(false)
+		case 23: // Not italic, not Fraktur
+			pen.Italic(false)
+		case 24: // Not underlined
+			pen.Underline(false)
+		case 25: // Blink off
+			pen.SlowBlink(false).RapidBlink(false)
+		case 27: // Positive (not reverse)
+			pen.Reverse(false)
+		case 28: // Reveal
+			pen.Conceal(false)
+		case 29: // Not crossed out
+			pen.Strikethrough(false)
+		case 30, 31, 32, 33, 34, 35, 36, 37: // Set foreground
+			pen.Foreground(ansi.Black + ansi.BasicColor(param-30)) //nolint:gosec
+		case 38: // Set foreground 256 or truecolor
+			var c color.Color
+			n := ReadStyleColor(params[i:], &c)
+			if n > 0 {
+				pen.Foreground(c)
+				i += n - 1
+			}
+		case 39: // Default foreground
+			pen.Foreground(nil)
+		case 40, 41, 42, 43, 44, 45, 46, 47: // Set background
+			pen.Background(ansi.Black + ansi.BasicColor(param-40)) //nolint:gosec
+		case 48: // Set background 256 or truecolor
+			var c color.Color
+			n := ReadStyleColor(params[i:], &c)
+			if n > 0 {
+				pen.Background(c)
+				i += n - 1
+			}
+		case 49: // Default Background
+			pen.Background(nil)
+		case 58: // Set underline color
+			var c color.Color
+			n := ReadStyleColor(params[i:], &c)
+			if n > 0 {
+				pen.UnderlineColor(c)
+				i += n - 1
+			}
+		case 59: // Default underline color
+			pen.UnderlineColor(nil)
+		case 90, 91, 92, 93, 94, 95, 96, 97: // Set bright foreground
+			pen.Foreground(ansi.BrightBlack + ansi.BasicColor(param-90)) //nolint:gosec
+		case 100, 101, 102, 103, 104, 105, 106, 107: // Set bright background
+			pen.Background(ansi.BrightBlack + ansi.BasicColor(param-100)) //nolint:gosec
+		}
+	}
+}
+
+// ReadLink reads a hyperlink escape sequence from a data buffer.
+func ReadLink(p []byte, link *Link) {
+	params := bytes.Split(p, []byte{';'})
+	if len(params) != 3 {
+		return
+	}
+	link.Params = string(params[1])
+	link.URL = string(params[2])
+}
+
+// ReadStyleColor reads a color from a list of parameters.
+// See [ansi.ReadStyleColor] for more information.
+func ReadStyleColor(params ansi.Params, c *color.Color) int {
+	return ansi.ReadStyleColor(params, c)
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/style.go b/vendor/github.com/charmbracelet/x/cellbuf/style.go
new file mode 100644
index 0000000..82c4afb
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/style.go
@@ -0,0 +1,31 @@
+package cellbuf
+
+import (
+	"github.com/charmbracelet/colorprofile"
+)
+
+// Convert converts a style to respect the given color profile.
+func ConvertStyle(s Style, p colorprofile.Profile) Style {
+	switch p {
+	case colorprofile.TrueColor:
+		return s
+	case colorprofile.Ascii:
+		s.Fg = nil
+		s.Bg = nil
+		s.Ul = nil
+	case colorprofile.NoTTY:
+		return Style{}
+	}
+
+	if s.Fg != nil {
+		s.Fg = p.Convert(s.Fg)
+	}
+	if s.Bg != nil {
+		s.Bg = p.Convert(s.Bg)
+	}
+	if s.Ul != nil {
+		s.Ul = p.Convert(s.Ul)
+	}
+
+	return s
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/tabstop.go b/vendor/github.com/charmbracelet/x/cellbuf/tabstop.go
new file mode 100644
index 0000000..24eec44
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/tabstop.go
@@ -0,0 +1,137 @@
+package cellbuf
+
+// DefaultTabInterval is the default tab interval.
+const DefaultTabInterval = 8
+
+// TabStops represents horizontal line tab stops.
+type TabStops struct {
+	stops    []int
+	interval int
+	width    int
+}
+
+// NewTabStops creates a new set of tab stops from a number of columns and an
+// interval.
+func NewTabStops(width, interval int) *TabStops {
+	ts := new(TabStops)
+	ts.interval = interval
+	ts.width = width
+	ts.stops = make([]int, (width+(interval-1))/interval)
+	ts.init(0, width)
+	return ts
+}
+
+// DefaultTabStops creates a new set of tab stops with the default interval.
+func DefaultTabStops(cols int) *TabStops {
+	return NewTabStops(cols, DefaultTabInterval)
+}
+
+// Resize resizes the tab stops to the given width.
+func (ts *TabStops) Resize(width int) {
+	if width == ts.width {
+		return
+	}
+
+	if width < ts.width {
+		size := (width + (ts.interval - 1)) / ts.interval
+		ts.stops = ts.stops[:size]
+	} else {
+		size := (width - ts.width + (ts.interval - 1)) / ts.interval
+		ts.stops = append(ts.stops, make([]int, size)...)
+	}
+
+	ts.init(ts.width, width)
+	ts.width = width
+}
+
+// IsStop returns true if the given column is a tab stop.
+func (ts TabStops) IsStop(col int) bool {
+	mask := ts.mask(col)
+	i := col >> 3
+	if i < 0 || i >= len(ts.stops) {
+		return false
+	}
+	return ts.stops[i]&mask != 0
+}
+
+// Next returns the next tab stop after the given column.
+func (ts TabStops) Next(col int) int {
+	return ts.Find(col, 1)
+}
+
+// Prev returns the previous tab stop before the given column.
+func (ts TabStops) Prev(col int) int {
+	return ts.Find(col, -1)
+}
+
+// Find returns the prev/next tab stop before/after the given column and delta.
+// If delta is positive, it returns the next tab stop after the given column.
+// If delta is negative, it returns the previous tab stop before the given column.
+// If delta is zero, it returns the given column.
+func (ts TabStops) Find(col, delta int) int {
+	if delta == 0 {
+		return col
+	}
+
+	var prev bool
+	count := delta
+	if count < 0 {
+		count = -count
+		prev = true
+	}
+
+	for count > 0 {
+		if !prev {
+			if col >= ts.width-1 {
+				return col
+			}
+
+			col++
+		} else {
+			if col < 1 {
+				return col
+			}
+
+			col--
+		}
+
+		if ts.IsStop(col) {
+			count--
+		}
+	}
+
+	return col
+}
+
+// Set adds a tab stop at the given column.
+func (ts *TabStops) Set(col int) {
+	mask := ts.mask(col)
+	ts.stops[col>>3] |= mask
+}
+
+// Reset removes the tab stop at the given column.
+func (ts *TabStops) Reset(col int) {
+	mask := ts.mask(col)
+	ts.stops[col>>3] &= ^mask
+}
+
+// Clear removes all tab stops.
+func (ts *TabStops) Clear() {
+	ts.stops = make([]int, len(ts.stops))
+}
+
+// mask returns the mask for the given column.
+func (ts *TabStops) mask(col int) int {
+	return 1 << (col & (ts.interval - 1))
+}
+
+// init initializes the tab stops starting from col until width.
+func (ts *TabStops) init(col, width int) {
+	for x := col; x < width; x++ {
+		if x%ts.interval == 0 {
+			ts.Set(x)
+		} else {
+			ts.Reset(x)
+		}
+	}
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/utils.go b/vendor/github.com/charmbracelet/x/cellbuf/utils.go
new file mode 100644
index 0000000..b0452fa
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/utils.go
@@ -0,0 +1,38 @@
+package cellbuf
+
+import (
+	"strings"
+)
+
+// Height returns the height of a string.
+func Height(s string) int {
+	return strings.Count(s, "\n") + 1
+}
+
+func min(a, b int) int { //nolint:predeclared
+	if a > b {
+		return b
+	}
+	return a
+}
+
+func max(a, b int) int { //nolint:predeclared
+	if a > b {
+		return a
+	}
+	return b
+}
+
+func clamp(v, low, high int) int {
+	if high < low {
+		low, high = high, low
+	}
+	return min(high, max(low, v))
+}
+
+func abs(a int) int {
+	if a < 0 {
+		return -a
+	}
+	return a
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/wrap.go b/vendor/github.com/charmbracelet/x/cellbuf/wrap.go
new file mode 100644
index 0000000..59a2a33
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/wrap.go
@@ -0,0 +1,178 @@
+package cellbuf
+
+import (
+	"bytes"
+	"unicode"
+	"unicode/utf8"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// Wrap returns a string that is wrapped to the specified limit applying any
+// ANSI escape sequences in the string. It tries to wrap the string at word
+// boundaries, but will break words if necessary.
+//
+// The breakpoints string is a list of characters that are considered
+// breakpoints for word wrapping. A hyphen (-) is always considered a
+// breakpoint.
+//
+// Note: breakpoints must be a string of 1-cell wide rune characters.
+func Wrap(s string, limit int, breakpoints string) string {
+	if len(s) == 0 {
+		return ""
+	}
+
+	if limit < 1 {
+		return s
+	}
+
+	p := ansi.GetParser()
+	defer ansi.PutParser(p)
+
+	var (
+		buf             bytes.Buffer
+		word            bytes.Buffer
+		space           bytes.Buffer
+		style, curStyle Style
+		link, curLink   Link
+		curWidth        int
+		wordLen         int
+	)
+
+	addSpace := func() {
+		curWidth += space.Len()
+		buf.Write(space.Bytes())
+		space.Reset()
+	}
+
+	addWord := func() {
+		if word.Len() == 0 {
+			return
+		}
+
+		curLink = link
+		curStyle = style
+
+		addSpace()
+		curWidth += wordLen
+		buf.Write(word.Bytes())
+		word.Reset()
+		wordLen = 0
+	}
+
+	addNewline := func() {
+		if !curStyle.Empty() {
+			buf.WriteString(ansi.ResetStyle)
+		}
+		if !curLink.Empty() {
+			buf.WriteString(ansi.ResetHyperlink())
+		}
+		buf.WriteByte('\n')
+		if !curLink.Empty() {
+			buf.WriteString(ansi.SetHyperlink(curLink.URL, curLink.Params))
+		}
+		if !curStyle.Empty() {
+			buf.WriteString(curStyle.Sequence())
+		}
+		curWidth = 0
+		space.Reset()
+	}
+
+	var state byte
+	for len(s) > 0 {
+		seq, width, n, newState := ansi.DecodeSequence(s, state, p)
+		switch width {
+		case 0:
+			if ansi.Equal(seq, "\t") {
+				addWord()
+				space.WriteString(seq)
+				break
+			} else if ansi.Equal(seq, "\n") {
+				if wordLen == 0 {
+					if curWidth+space.Len() > limit {
+						curWidth = 0
+					} else {
+						// preserve whitespaces
+						buf.Write(space.Bytes())
+					}
+					space.Reset()
+				}
+
+				addWord()
+				addNewline()
+				break
+			} else if ansi.HasCsiPrefix(seq) && p.Command() == 'm' {
+				// SGR style sequence [ansi.SGR]
+				ReadStyle(p.Params(), &style)
+			} else if ansi.HasOscPrefix(seq) && p.Command() == 8 {
+				// Hyperlink sequence [ansi.SetHyperlink]
+				ReadLink(p.Data(), &link)
+			}
+
+			word.WriteString(seq)
+		default:
+			if len(seq) == 1 {
+				// ASCII
+				r, _ := utf8.DecodeRuneInString(seq)
+				if unicode.IsSpace(r) {
+					addWord()
+					space.WriteRune(r)
+					break
+				} else if r == '-' || runeContainsAny(r, breakpoints) {
+					addSpace()
+					if curWidth+wordLen+width <= limit {
+						addWord()
+						buf.WriteString(seq)
+						curWidth += width
+						break
+					}
+				}
+			}
+
+			if wordLen+width > limit {
+				// Hardwrap the word if it's too long
+				addWord()
+			}
+
+			word.WriteString(seq)
+			wordLen += width
+
+			if curWidth+wordLen+space.Len() > limit {
+				addNewline()
+			}
+		}
+
+		s = s[n:]
+		state = newState
+	}
+
+	if wordLen == 0 {
+		if curWidth+space.Len() > limit {
+			curWidth = 0
+		} else {
+			// preserve whitespaces
+			buf.Write(space.Bytes())
+		}
+		space.Reset()
+	}
+
+	addWord()
+
+	if !curLink.Empty() {
+		buf.WriteString(ansi.ResetHyperlink())
+	}
+	if !curStyle.Empty() {
+		buf.WriteString(ansi.ResetStyle)
+	}
+
+	return buf.String()
+}
+
+func runeContainsAny[T string | []rune](r rune, s T) bool {
+	for _, c := range []rune(s) {
+		if c == r {
+			return true
+		}
+	}
+	return false
+}
diff --git a/vendor/github.com/charmbracelet/x/cellbuf/writer.go b/vendor/github.com/charmbracelet/x/cellbuf/writer.go
new file mode 100644
index 0000000..ae8b2a8
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/cellbuf/writer.go
@@ -0,0 +1,339 @@
+package cellbuf
+
+import (
+	"bytes"
+	"fmt"
+	"strings"
+
+	"github.com/charmbracelet/x/ansi"
+)
+
+// CellBuffer is a cell buffer that represents a set of cells in a screen or a
+// grid.
+type CellBuffer interface {
+	// Cell returns the cell at the given position.
+	Cell(x, y int) *Cell
+	// SetCell sets the cell at the given position to the given cell. It
+	// returns whether the cell was set successfully.
+	SetCell(x, y int, c *Cell) bool
+	// Bounds returns the bounds of the cell buffer.
+	Bounds() Rectangle
+}
+
+// FillRect fills the rectangle within the cell buffer with the given cell.
+// This will not fill cells outside the bounds of the cell buffer.
+func FillRect(s CellBuffer, c *Cell, rect Rectangle) {
+	for y := rect.Min.Y; y < rect.Max.Y; y++ {
+		for x := rect.Min.X; x < rect.Max.X; x++ {
+			s.SetCell(x, y, c) //nolint:errcheck
+		}
+	}
+}
+
+// Fill fills the cell buffer with the given cell.
+func Fill(s CellBuffer, c *Cell) {
+	FillRect(s, c, s.Bounds())
+}
+
+// ClearRect clears the rectangle within the cell buffer with blank cells.
+func ClearRect(s CellBuffer, rect Rectangle) {
+	FillRect(s, nil, rect)
+}
+
+// Clear clears the cell buffer with blank cells.
+func Clear(s CellBuffer) {
+	Fill(s, nil)
+}
+
+// SetContentRect clears the rectangle within the cell buffer with blank cells,
+// and sets the given string as its content. If the height or width of the
+// string exceeds the height or width of the cell buffer, it will be truncated.
+func SetContentRect(s CellBuffer, str string, rect Rectangle) {
+	// Replace all "\n" with "\r\n" to ensure the cursor is reset to the start
+	// of the line. Make sure we don't replace "\r\n" with "\r\r\n".
+	str = strings.ReplaceAll(str, "\r\n", "\n")
+	str = strings.ReplaceAll(str, "\n", "\r\n")
+	ClearRect(s, rect)
+	printString(s, ansi.GraphemeWidth, rect.Min.X, rect.Min.Y, rect, str, true, "")
+}
+
+// SetContent clears the cell buffer with blank cells, and sets the given string
+// as its content. If the height or width of the string exceeds the height or
+// width of the cell buffer, it will be truncated.
+func SetContent(s CellBuffer, str string) {
+	SetContentRect(s, str, s.Bounds())
+}
+
+// Render returns a string representation of the grid with ANSI escape sequences.
+func Render(d CellBuffer) string {
+	var buf bytes.Buffer
+	height := d.Bounds().Dy()
+	for y := 0; y < height; y++ {
+		_, line := RenderLine(d, y)
+		buf.WriteString(line)
+		if y < height-1 {
+			buf.WriteString("\r\n")
+		}
+	}
+	return buf.String()
+}
+
+// RenderLine returns a string representation of the yth line of the grid along
+// with the width of the line.
+func RenderLine(d CellBuffer, n int) (w int, line string) {
+	var pen Style
+	var link Link
+	var buf bytes.Buffer
+	var pendingLine string
+	var pendingWidth int // this ignores space cells until we hit a non-space cell
+
+	writePending := func() {
+		// If there's no pending line, we don't need to do anything.
+		if len(pendingLine) == 0 {
+			return
+		}
+		buf.WriteString(pendingLine)
+		w += pendingWidth
+		pendingWidth = 0
+		pendingLine = ""
+	}
+
+	for x := 0; x < d.Bounds().Dx(); x++ {
+		if cell := d.Cell(x, n); cell != nil && cell.Width > 0 {
+			// Convert the cell's style and link to the given color profile.
+			cellStyle := cell.Style
+			cellLink := cell.Link
+			if cellStyle.Empty() && !pen.Empty() {
+				writePending()
+				buf.WriteString(ansi.ResetStyle) //nolint:errcheck
+				pen.Reset()
+			}
+			if !cellStyle.Equal(&pen) {
+				writePending()
+				seq := cellStyle.DiffSequence(pen)
+				buf.WriteString(seq) // nolint:errcheck
+				pen = cellStyle
+			}
+
+			// Write the URL escape sequence
+			if cellLink != link && link.URL != "" {
+				writePending()
+				buf.WriteString(ansi.ResetHyperlink()) //nolint:errcheck
+				link.Reset()
+			}
+			if cellLink != link {
+				writePending()
+				buf.WriteString(ansi.SetHyperlink(cellLink.URL, cellLink.Params)) //nolint:errcheck
+				link = cellLink
+			}
+
+			// We only write the cell content if it's not empty. If it is, we
+			// append it to the pending line and width to be evaluated later.
+			if cell.Equal(&BlankCell) {
+				pendingLine += cell.String()
+				pendingWidth += cell.Width
+			} else {
+				writePending()
+				buf.WriteString(cell.String())
+				w += cell.Width
+			}
+		}
+	}
+	if link.URL != "" {
+		buf.WriteString(ansi.ResetHyperlink()) //nolint:errcheck
+	}
+	if !pen.Empty() {
+		buf.WriteString(ansi.ResetStyle) //nolint:errcheck
+	}
+	return w, strings.TrimRight(buf.String(), " ") // Trim trailing spaces
+}
+
+// ScreenWriter represents a writer that writes to a [Screen] parsing ANSI
+// escape sequences and Unicode characters and converting them into cells that
+// can be written to a cell [Buffer].
+type ScreenWriter struct {
+	*Screen
+}
+
+// NewScreenWriter creates a new ScreenWriter that writes to the given Screen.
+// This is a convenience function for creating a ScreenWriter.
+func NewScreenWriter(s *Screen) *ScreenWriter {
+	return &ScreenWriter{s}
+}
+
+// Write writes the given bytes to the screen.
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape
+// sequences.
+func (s *ScreenWriter) Write(p []byte) (n int, err error) {
+	printString(s.Screen, s.method,
+		s.cur.X, s.cur.Y, s.Bounds(),
+		p, false, "")
+	return len(p), nil
+}
+
+// SetContent clears the screen with blank cells, and sets the given string as
+// its content. If the height or width of the string exceeds the height or
+// width of the screen, it will be truncated.
+//
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape sequences.
+func (s *ScreenWriter) SetContent(str string) {
+	s.SetContentRect(str, s.Bounds())
+}
+
+// SetContentRect clears the rectangle within the screen with blank cells, and
+// sets the given string as its content. If the height or width of the string
+// exceeds the height or width of the screen, it will be truncated.
+//
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape
+// sequences.
+func (s *ScreenWriter) SetContentRect(str string, rect Rectangle) {
+	// Replace all "\n" with "\r\n" to ensure the cursor is reset to the start
+	// of the line. Make sure we don't replace "\r\n" with "\r\r\n".
+	str = strings.ReplaceAll(str, "\r\n", "\n")
+	str = strings.ReplaceAll(str, "\n", "\r\n")
+	s.ClearRect(rect)
+	printString(s.Screen, s.method,
+		rect.Min.X, rect.Min.Y, rect,
+		str, true, "")
+}
+
+// Print prints the string at the current cursor position. It will wrap the
+// string to the width of the screen if it exceeds the width of the screen.
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape
+// sequences.
+func (s *ScreenWriter) Print(str string, v ...interface{}) {
+	if len(v) > 0 {
+		str = fmt.Sprintf(str, v...)
+	}
+	printString(s.Screen, s.method,
+		s.cur.X, s.cur.Y, s.Bounds(),
+		str, false, "")
+}
+
+// PrintAt prints the string at the given position. It will wrap the string to
+// the width of the screen if it exceeds the width of the screen.
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape
+// sequences.
+func (s *ScreenWriter) PrintAt(x, y int, str string, v ...interface{}) {
+	if len(v) > 0 {
+		str = fmt.Sprintf(str, v...)
+	}
+	printString(s.Screen, s.method,
+		x, y, s.Bounds(),
+		str, false, "")
+}
+
+// PrintCrop prints the string at the current cursor position and truncates the
+// text if it exceeds the width of the screen. Use tail to specify a string to
+// append if the string is truncated.
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape
+// sequences.
+func (s *ScreenWriter) PrintCrop(str string, tail string) {
+	printString(s.Screen, s.method,
+		s.cur.X, s.cur.Y, s.Bounds(),
+		str, true, tail)
+}
+
+// PrintCropAt prints the string at the given position and truncates the text
+// if it exceeds the width of the screen. Use tail to specify a string to append
+// if the string is truncated.
+// This will recognize ANSI [ansi.SGR] style and [ansi.SetHyperlink] escape
+// sequences.
+func (s *ScreenWriter) PrintCropAt(x, y int, str string, tail string) {
+	printString(s.Screen, s.method,
+		x, y, s.Bounds(),
+		str, true, tail)
+}
+
+// printString draws a string starting at the given position.
+func printString[T []byte | string](
+	s CellBuffer,
+	m ansi.Method,
+	x, y int,
+	bounds Rectangle, str T,
+	truncate bool, tail string,
+) {
+	p := ansi.GetParser()
+	defer ansi.PutParser(p)
+
+	var tailc Cell
+	if truncate && len(tail) > 0 {
+		if m == ansi.WcWidth {
+			tailc = *NewCellString(tail)
+		} else {
+			tailc = *NewGraphemeCell(tail)
+		}
+	}
+
+	decoder := ansi.DecodeSequenceWc[T]
+	if m == ansi.GraphemeWidth {
+		decoder = ansi.DecodeSequence[T]
+	}
+
+	var cell Cell
+	var style Style
+	var link Link
+	var state byte
+	for len(str) > 0 {
+		seq, width, n, newState := decoder(str, state, p)
+
+		switch width {
+		case 1, 2, 3, 4: // wide cells can go up to 4 cells wide
+			cell.Width += width
+			cell.Append([]rune(string(seq))...)
+
+			if !truncate && x+cell.Width > bounds.Max.X && y+1 < bounds.Max.Y {
+				// Wrap the string to the width of the window
+				x = bounds.Min.X
+				y++
+			}
+			if Pos(x, y).In(bounds) {
+				if truncate && tailc.Width > 0 && x+cell.Width > bounds.Max.X-tailc.Width {
+					// Truncate the string and append the tail if any.
+					cell := tailc
+					cell.Style = style
+					cell.Link = link
+					s.SetCell(x, y, &cell)
+					x += tailc.Width
+				} else {
+					// Print the cell to the screen
+					cell.Style = style
+					cell.Link = link
+					s.SetCell(x, y, &cell) //nolint:errcheck
+					x += width
+				}
+			}
+
+			// String is too long for the line, truncate it.
+			// Make sure we reset the cell for the next iteration.
+			cell.Reset()
+		default:
+			// Valid sequences always have a non-zero Cmd.
+			// TODO: Handle cursor movement and other sequences
+			switch {
+			case ansi.HasCsiPrefix(seq) && p.Command() == 'm':
+				// SGR - Select Graphic Rendition
+				ReadStyle(p.Params(), &style)
+			case ansi.HasOscPrefix(seq) && p.Command() == 8:
+				// Hyperlinks
+				ReadLink(p.Data(), &link)
+			case ansi.Equal(seq, T("\n")):
+				y++
+			case ansi.Equal(seq, T("\r")):
+				x = bounds.Min.X
+			default:
+				cell.Append([]rune(string(seq))...)
+			}
+		}
+
+		// Advance the state and data
+		state = newState
+		str = str[n:]
+	}
+
+	// Make sure to set the last cell if it's not empty.
+	if !cell.Empty() {
+		s.SetCell(x, y, &cell) //nolint:errcheck
+		cell.Reset()
+	}
+}
diff --git a/vendor/github.com/charmbracelet/x/term/LICENSE b/vendor/github.com/charmbracelet/x/term/LICENSE
new file mode 100644
index 0000000..65a5654
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/LICENSE
@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2023 Charmbracelet, Inc.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
diff --git a/vendor/github.com/charmbracelet/x/term/term.go b/vendor/github.com/charmbracelet/x/term/term.go
new file mode 100644
index 0000000..58d6522
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/term.go
@@ -0,0 +1,49 @@
+package term
+
+// State contains platform-specific state of a terminal.
+type State struct {
+	state
+}
+
+// IsTerminal returns whether the given file descriptor is a terminal.
+func IsTerminal(fd uintptr) bool {
+	return isTerminal(fd)
+}
+
+// MakeRaw puts the terminal connected to the given file descriptor into raw
+// mode and returns the previous state of the terminal so that it can be
+// restored.
+func MakeRaw(fd uintptr) (*State, error) {
+	return makeRaw(fd)
+}
+
+// GetState returns the current state of a terminal which may be useful to
+// restore the terminal after a signal.
+func GetState(fd uintptr) (*State, error) {
+	return getState(fd)
+}
+
+// SetState sets the given state of the terminal.
+func SetState(fd uintptr, state *State) error {
+	return setState(fd, state)
+}
+
+// Restore restores the terminal connected to the given file descriptor to a
+// previous state.
+func Restore(fd uintptr, oldState *State) error {
+	return restore(fd, oldState)
+}
+
+// GetSize returns the visible dimensions of the given terminal.
+//
+// These dimensions don't include any scrollback buffer height.
+func GetSize(fd uintptr) (width, height int, err error) {
+	return getSize(fd)
+}
+
+// ReadPassword reads a line of input from a terminal without local echo.  This
+// is commonly used for inputting passwords and other sensitive data. The slice
+// returned does not include the \n.
+func ReadPassword(fd uintptr) ([]byte, error) {
+	return readPassword(fd)
+}
diff --git a/vendor/github.com/charmbracelet/x/term/term_other.go b/vendor/github.com/charmbracelet/x/term/term_other.go
new file mode 100644
index 0000000..092c7e9
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/term_other.go
@@ -0,0 +1,39 @@
+//go:build !aix && !darwin && !dragonfly && !freebsd && !linux && !netbsd && !openbsd && !zos && !windows && !solaris && !plan9
+// +build !aix,!darwin,!dragonfly,!freebsd,!linux,!netbsd,!openbsd,!zos,!windows,!solaris,!plan9
+
+package term
+
+import (
+	"fmt"
+	"runtime"
+)
+
+type state struct{}
+
+func isTerminal(fd uintptr) bool {
+	return false
+}
+
+func makeRaw(fd uintptr) (*State, error) {
+	return nil, fmt.Errorf("terminal: MakeRaw not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
+}
+
+func getState(fd uintptr) (*State, error) {
+	return nil, fmt.Errorf("terminal: GetState not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
+}
+
+func restore(fd uintptr, state *State) error {
+	return fmt.Errorf("terminal: Restore not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
+}
+
+func getSize(fd uintptr) (width, height int, err error) {
+	return 0, 0, fmt.Errorf("terminal: GetSize not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
+}
+
+func setState(fd uintptr, state *State) error {
+	return fmt.Errorf("terminal: SetState not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
+}
+
+func readPassword(fd uintptr) ([]byte, error) {
+	return nil, fmt.Errorf("terminal: ReadPassword not implemented on %s/%s", runtime.GOOS, runtime.GOARCH)
+}
diff --git a/vendor/github.com/charmbracelet/x/term/term_unix.go b/vendor/github.com/charmbracelet/x/term/term_unix.go
new file mode 100644
index 0000000..1459cb1
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/term_unix.go
@@ -0,0 +1,96 @@
+//go:build aix || darwin || dragonfly || freebsd || linux || netbsd || openbsd || solaris || zos
+// +build aix darwin dragonfly freebsd linux netbsd openbsd solaris zos
+
+package term
+
+import (
+	"golang.org/x/sys/unix"
+)
+
+type state struct {
+	unix.Termios
+}
+
+func isTerminal(fd uintptr) bool {
+	_, err := unix.IoctlGetTermios(int(fd), ioctlReadTermios)
+	return err == nil
+}
+
+func makeRaw(fd uintptr) (*State, error) {
+	termios, err := unix.IoctlGetTermios(int(fd), ioctlReadTermios)
+	if err != nil {
+		return nil, err
+	}
+
+	oldState := State{state{Termios: *termios}}
+
+	// This attempts to replicate the behaviour documented for cfmakeraw in
+	// the termios(3) manpage.
+	termios.Iflag &^= unix.IGNBRK | unix.BRKINT | unix.PARMRK | unix.ISTRIP | unix.INLCR | unix.IGNCR | unix.ICRNL | unix.IXON
+	termios.Oflag &^= unix.OPOST
+	termios.Lflag &^= unix.ECHO | unix.ECHONL | unix.ICANON | unix.ISIG | unix.IEXTEN
+	termios.Cflag &^= unix.CSIZE | unix.PARENB
+	termios.Cflag |= unix.CS8
+	termios.Cc[unix.VMIN] = 1
+	termios.Cc[unix.VTIME] = 0
+	if err := unix.IoctlSetTermios(int(fd), ioctlWriteTermios, termios); err != nil {
+		return nil, err
+	}
+
+	return &oldState, nil
+}
+
+func setState(fd uintptr, state *State) error {
+	var termios *unix.Termios
+	if state != nil {
+		termios = &state.Termios
+	}
+	return unix.IoctlSetTermios(int(fd), ioctlWriteTermios, termios)
+}
+
+func getState(fd uintptr) (*State, error) {
+	termios, err := unix.IoctlGetTermios(int(fd), ioctlReadTermios)
+	if err != nil {
+		return nil, err
+	}
+
+	return &State{state{Termios: *termios}}, nil
+}
+
+func restore(fd uintptr, state *State) error {
+	return unix.IoctlSetTermios(int(fd), ioctlWriteTermios, &state.Termios)
+}
+
+func getSize(fd uintptr) (width, height int, err error) {
+	ws, err := unix.IoctlGetWinsize(int(fd), unix.TIOCGWINSZ)
+	if err != nil {
+		return 0, 0, err
+	}
+	return int(ws.Col), int(ws.Row), nil
+}
+
+// passwordReader is an io.Reader that reads from a specific file descriptor.
+type passwordReader int
+
+func (r passwordReader) Read(buf []byte) (int, error) {
+	return unix.Read(int(r), buf)
+}
+
+func readPassword(fd uintptr) ([]byte, error) {
+	termios, err := unix.IoctlGetTermios(int(fd), ioctlReadTermios)
+	if err != nil {
+		return nil, err
+	}
+
+	newState := *termios
+	newState.Lflag &^= unix.ECHO
+	newState.Lflag |= unix.ICANON | unix.ISIG
+	newState.Iflag |= unix.ICRNL
+	if err := unix.IoctlSetTermios(int(fd), ioctlWriteTermios, &newState); err != nil {
+		return nil, err
+	}
+
+	defer unix.IoctlSetTermios(int(fd), ioctlWriteTermios, termios)
+
+	return readPasswordLine(passwordReader(fd))
+}
diff --git a/vendor/github.com/charmbracelet/x/term/term_unix_bsd.go b/vendor/github.com/charmbracelet/x/term/term_unix_bsd.go
new file mode 100644
index 0000000..b435031
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/term_unix_bsd.go
@@ -0,0 +1,11 @@
+//go:build darwin || dragonfly || freebsd || netbsd || openbsd
+// +build darwin dragonfly freebsd netbsd openbsd
+
+package term
+
+import "golang.org/x/sys/unix"
+
+const (
+	ioctlReadTermios  = unix.TIOCGETA
+	ioctlWriteTermios = unix.TIOCSETA
+)
diff --git a/vendor/github.com/charmbracelet/x/term/term_unix_other.go b/vendor/github.com/charmbracelet/x/term/term_unix_other.go
new file mode 100644
index 0000000..ee2a29e
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/term_unix_other.go
@@ -0,0 +1,11 @@
+//go:build aix || linux || solaris || zos
+// +build aix linux solaris zos
+
+package term
+
+import "golang.org/x/sys/unix"
+
+const (
+	ioctlReadTermios  = unix.TCGETS
+	ioctlWriteTermios = unix.TCSETS
+)
diff --git a/vendor/github.com/charmbracelet/x/term/term_windows.go b/vendor/github.com/charmbracelet/x/term/term_windows.go
new file mode 100644
index 0000000..fe7afde
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/term_windows.go
@@ -0,0 +1,87 @@
+//go:build windows
+// +build windows
+
+package term
+
+import (
+	"os"
+
+	"golang.org/x/sys/windows"
+)
+
+type state struct {
+	Mode uint32
+}
+
+func isTerminal(fd uintptr) bool {
+	var st uint32
+	err := windows.GetConsoleMode(windows.Handle(fd), &st)
+	return err == nil
+}
+
+func makeRaw(fd uintptr) (*State, error) {
+	var st uint32
+	if err := windows.GetConsoleMode(windows.Handle(fd), &st); err != nil {
+		return nil, err
+	}
+	raw := st &^ (windows.ENABLE_ECHO_INPUT | windows.ENABLE_PROCESSED_INPUT | windows.ENABLE_LINE_INPUT | windows.ENABLE_PROCESSED_OUTPUT)
+	raw |= windows.ENABLE_VIRTUAL_TERMINAL_INPUT
+	if err := windows.SetConsoleMode(windows.Handle(fd), raw); err != nil {
+		return nil, err
+	}
+	return &State{state{st}}, nil
+}
+
+func setState(fd uintptr, state *State) error {
+	var mode uint32
+	if state != nil {
+		mode = state.Mode
+	}
+	return windows.SetConsoleMode(windows.Handle(fd), mode)
+}
+
+func getState(fd uintptr) (*State, error) {
+	var st uint32
+	if err := windows.GetConsoleMode(windows.Handle(fd), &st); err != nil {
+		return nil, err
+	}
+	return &State{state{st}}, nil
+}
+
+func restore(fd uintptr, state *State) error {
+	return windows.SetConsoleMode(windows.Handle(fd), state.Mode)
+}
+
+func getSize(fd uintptr) (width, height int, err error) {
+	var info windows.ConsoleScreenBufferInfo
+	if err := windows.GetConsoleScreenBufferInfo(windows.Handle(fd), &info); err != nil {
+		return 0, 0, err
+	}
+	return int(info.Window.Right - info.Window.Left + 1), int(info.Window.Bottom - info.Window.Top + 1), nil
+}
+
+func readPassword(fd uintptr) ([]byte, error) {
+	var st uint32
+	if err := windows.GetConsoleMode(windows.Handle(fd), &st); err != nil {
+		return nil, err
+	}
+	old := st
+
+	st &^= (windows.ENABLE_ECHO_INPUT | windows.ENABLE_LINE_INPUT)
+	st |= (windows.ENABLE_PROCESSED_OUTPUT | windows.ENABLE_PROCESSED_INPUT)
+	if err := windows.SetConsoleMode(windows.Handle(fd), st); err != nil {
+		return nil, err
+	}
+
+	defer windows.SetConsoleMode(windows.Handle(fd), old)
+
+	var h windows.Handle
+	p, _ := windows.GetCurrentProcess()
+	if err := windows.DuplicateHandle(p, windows.Handle(fd), p, &h, 0, false, windows.DUPLICATE_SAME_ACCESS); err != nil {
+		return nil, err
+	}
+
+	f := os.NewFile(uintptr(h), "stdin")
+	defer f.Close()
+	return readPasswordLine(f)
+}
diff --git a/vendor/github.com/charmbracelet/x/term/terminal.go b/vendor/github.com/charmbracelet/x/term/terminal.go
new file mode 100644
index 0000000..8963163
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/terminal.go
@@ -0,0 +1,12 @@
+package term
+
+import (
+	"io"
+)
+
+// File represents a file that has a file descriptor and can be read from,
+// written to, and closed.
+type File interface {
+	io.ReadWriteCloser
+	Fd() uintptr
+}
diff --git a/vendor/github.com/charmbracelet/x/term/util.go b/vendor/github.com/charmbracelet/x/term/util.go
new file mode 100644
index 0000000..b731341
--- /dev/null
+++ b/vendor/github.com/charmbracelet/x/term/util.go
@@ -0,0 +1,47 @@
+package term
+
+import (
+	"io"
+	"runtime"
+)
+
+// readPasswordLine reads from reader until it finds \n or io.EOF.
+// The slice returned does not include the \n.
+// readPasswordLine also ignores any \r it finds.
+// Windows uses \r as end of line. So, on Windows, readPasswordLine
+// reads until it finds \r and ignores any \n it finds during processing.
+func readPasswordLine(reader io.Reader) ([]byte, error) {
+	var buf [1]byte
+	var ret []byte
+
+	for {
+		n, err := reader.Read(buf[:])
+		if n > 0 {
+			switch buf[0] {
+			case '\b':
+				if len(ret) > 0 {
+					ret = ret[:len(ret)-1]
+				}
+			case '\n':
+				if runtime.GOOS != "windows" {
+					return ret, nil
+				}
+				// otherwise ignore \n
+			case '\r':
+				if runtime.GOOS == "windows" {
+					return ret, nil
+				}
+				// otherwise ignore \r
+			default:
+				ret = append(ret, buf[0])
+			}
+			continue
+		}
+		if err != nil {
+			if err == io.EOF && len(ret) > 0 {
+				return ret, nil
+			}
+			return ret, err
+		}
+	}
+}