* fix(backend): strip non-ASCII chars from branch/worktree names
Non-ASCII titles (CJK, Cyrillic, Arabic, emoji) flowed verbatim into git
branch names and worktree directory names because the sanitizers used
unicode.IsLetter/IsDigit, which accept any Unicode letter or digit. The
resulting names broke many downstream tools and left worktrees
effectively inaccessible.
Restrict SanitizeForBranch and SanitizeRepoDirName to ASCII
alphanumerics. Non-ASCII runes now route through the existing
hyphen-replace + collapse + trim pipeline, and the existing fallback
chain (sanitized title -> sanitized task ID -> "task") handles
titles that become empty after stripping. The "Issue #N:" / "PR #N:"
prefix already added by the GitHub ingestion path means a non-ASCII
issue title still produces a recognizable name like "issue-123".
Closes#1081
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* refactor(worktree): fix comment accuracy and harden test assertion
- Remove "emoji" from isASCIIAlphaNum doc — emoji are not Unicode
letters (unicode.IsLetter returns false), so the prior code never
admitted bare emoji. Correct examples to CJK, Cyrillic, Arabic.
- Replace hard-coded len(dirName) != 8 with strings.Contains check
so the assertion stays correct if suffix length ever changes.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Kandev Agent <agent@kandev.dev>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
João Salavisa
7be0db04d8