triage
This skill uses Claude hooks which can execute code automatically in response to events. Review carefully before installing.
MANDATORY: Follow the complete shared lifecycle contract in ${CLAUDE_PLUGIN_ROOT}/skills/shared/run-lifecycle.md with declared skill triage.
Triage
MANDATORY: READ ENTIRE FILE before composing user-facing prose: ${CLAUDE_PLUGIN_ROOT}/skills/shared/readability-style.md.
Investigate an existing issue against immutable evidence, then make only the verified update that prepares an eligible issue for /design. /triage never edits repository files and never authors a larch:plan block.
Anti-halt continuation reminder. After every child Skill tool call (e.g., /block-issue, /issue) returns AND after every numbered-step Bash helper call, IMMEDIATELY continue with this skill's NEXT numbered step - do NOT end the turn on the child's cleanup output or helper stdout, and do NOT write a summary, handoff, status recap, or "returning to parent" message - those are halts in disguise. The rule is strictly subordinate to any explicit non-sequential control-flow directive in THIS file (e.g., bail to cleanup, skip dependency processing). A normal sequential proceed to Step N+1 instruction is the default continuation this rule reinforces, NOT an exception. → shared/subskill-invocation.md#anti-halt
Public contract
/triage <issue-number> [--repo OWNER/REPO] [--report-only]
- Accept exactly one positive issue number, at most one validated
--repo OWNER/REPO, and the optional boolean--report-only. Reject every other flag or positional argument before allocating scratch space. - Verdicts are exactly
valid,already-fixed,duplicate,invalid, andinconclusive. - Render the evidence, diagnosis, verdict, missing evidence, and fix outline before the terminal machine lines.
- Emit exactly these terminal keys: