Emit OSC 9;4 progress start/end sequences during agent streaming and
compaction so terminals (iTerm2, WezTerm, Windows Terminal, Kitty, etc.)
show activity in their tab bar.
Made with love for @lucasmeijer
Only use fd --full-path for path queries so plain fuzzy @ autocomplete results no longer depend on whether the cwd path contains the query text.
closes#2778
* feat(coding-agent,tui): support argument-hint frontmatter in prompt templates
Parse argument-hint from prompt template frontmatter and display it
in the autocomplete dropdown description, matching Claude Code's
convention for custom commands.
Frontmatter format:
---
description: Code review
argument-hint: "[file | #PR | PR-URL]"
---
The hint renders in the description column of the autocomplete list:
review [file | #PR | PR-URL] — Code review
Closes#2761
* docs(coding-agent,tui): add argument-hint documentation, tests, and built-in hints
- Document argument-hint frontmatter in prompt-templates.md with <required>/[optional] convention
- Add argument-hint to built-in prompts: pr, is, wr
- Expand tests: required/optional hints, missing hints, empty hints, special characters
- Add changelog entries for coding-agent and tui
OSC 8 hyperlinks landed in #3248, but detectCapabilities() returned
hyperlinks: true in the unknown-terminal fallback. Terminals that
silently swallow OSC 8 (most xterm-compatible hosts, tmux/screen
without passthrough) end up dropping the URL from rendered markdown
links entirely, since the fallback 'text (url)' rendering is skipped
whenever hyperlinks is true.
- Unknown terminals now default to hyperlinks: false.
- tmux and screen (TMUX env, TERM starting with tmux/screen) force
hyperlinks: false even when the outer terminal would otherwise
advertise OSC 8 support. Image protocols also left disabled.
- Added detectCapabilities tests covering the known-capable set and
the tmux/screen/unknown cases.
TerminalCapabilities already tracks hyperlinks: boolean and returns true
for Ghostty, Kitty, WezTerm, and iTerm2, but nothing generated OSC 8
sequences. This completes that stub.
Changes to packages/tui:
- terminal-image.ts: add hyperlink(text, url) and setCapabilities()
- index.ts: export hyperlink and setCapabilities
- utils.ts: extend AnsiCodeTracker to track active OSC 8 URLs
- process() now handles OSC 8 open/close sequences
- getActiveCodes() re-emits the OSC 8 open at each line start
- getLineEndReset() closes the OSC 8 hyperlink before each line break
This ensures hyperlinks wrap correctly across multiple lines.
- components/markdown.ts: link renderer uses hyperlink() when
getCapabilities().hyperlinks is true; falls back to (url) text
- Tests: new wrap-ansi tests for OSC 8 line-wrapping; terminal-image
tests for hyperlink(); markdown tests covering both code paths;
table-cell width test pinned to hyperlinks:false (checks raw columns)
closes#3239
Co-authored-by: AI (Pi/Claude Sonnet 4.6) <noreply@pi.dev>
Co-authored-by: Mario Zechner <badlogicgames@gmail.com>
This reverts 3929e0c181.
The Zellij-specific Kitty-query skip regresses Shift+Enter newline handling in Zellij. Restoring the previous behavior for now while we work on a safer fix that preserves both Alt and modified Enter handling.\n\nContext: https://github.com/badlogic/pi-mono/issues/3259
The quoteInlineStyleContext used an empty stylePrefix, so after inline
elements (links, bold, code) closed their ANSI codes inside a blockquote,
the quote color was not re-applied. Text following a link rendered in
default foreground instead of the quote theme color.
Headings applied styling (bold+cyan) as an outer wrapper around the
result of renderInlineTokens. When inline elements like codespan emitted
ANSI resets, the outer heading style was killed and only the default
text style was restored via stylePrefix.
Fix: pass a heading-specific InlineStyleContext into renderInlineTokens
so each text segment gets heading styling directly, and stylePrefix
restores heading style after inline elements.
fullRender wrote \x1b[3J\x1b[2J\x1b[H (clear scrollback, then clear
screen). But \x1b[2J pushes on-screen content into scrollback, undoing
the preceding \x1b[3J. Reorder to \x1b[2J\x1b[H\x1b[3J so the screen is
cleared first and scrollback is wiped last.
0x08 (BS) was treated as plain backspace alongside 0x7f (DEL). On
Windows Terminal, Backspace sends 0x7f and Ctrl+Backspace sends 0x08,
making ctrl+backspace bindings unreachable.
Now 0x08 matches ctrl+backspace (and ctrl+h, same byte) but not plain
backspace. 0x7f remains plain backspace.
fixes#2139
* feat(tui): treat paste markers as atomic segments in editor
Paste markers like `[paste #1 +123 lines]` are now treated as single
atomic units for cursor movement, word navigation, deletion, and
wrapping. Only markers with valid paste IDs (present in the editor's
pastes Map) are treated atomically.
* fix(tui): word-wrap oversized atomic segments in editor
