Issue Triage

${var} — Repo (owner/repo) to triage. If empty, triages all watched repos.

If ${var} is set, only triage that repo (owner/repo format).

Config

This skill reads repos from memory/watched-repos.md. If the file doesn't exist yet, create it or skip this skill.

# memory/watched-repos.md
- owner/repo
- another-owner/another-repo

Read memory/MEMORY.md and memory/watched-repos.md for repos to triage. Read the last 2 days of memory/logs/ to avoid re-triaging.

Steps:

1. For each repo in watched-repos.md, list issues created in the last 48h:
   bashCopy

   gh issue list -R owner/repo --state open --json number,title,body,labels,createdAt --limit 20
   

2. For each new issue without labels:
   • Read the issue body and classify:
     • bug — something broken
     • feature — new capability request
     • question — needs clarification
     • urgent — security, data loss, or blocking
     • good-first-issue — well-scoped, self-contained
   • Apply appropriate labels:
     bashCopy

     gh issue edit NUMBER -R owner/repo --add-label "label"
     

   • If labeled urgent, also post a comment acknowledging it:
     bashCopy

     gh issue comment NUMBER -R owner/repo --body "Triaged as urgent — flagging for immediate attention."
     

3. Send a summary via ./notify if any urgent issues found.
4. Log what you triaged to memory/logs/${today}.md. If no new issues, log "ISSUE_TRIAGE_OK" and end.