bug-triage
Triage bugs reported in chat/issues, search for duplicates, file or update GitHub issues with full context, and push fix PRs.
npx skills add Untrivial-ai/agent-orchestrator --skill bug-triage --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Triage bugs into well-structured GitHub issues on the upstream AgentWrapper/agent-orchestrator repo. It handles duplicates, creates or updates issues with full context, and can push fix PRs.
How it works
- Pre-flight: Pull latest code from upstream, target repo is AgentWrapper/agent-orchestrator, verify ao status shows port 3001, and record the source data.
- Gather Context: Extract reporter, description, attachments, and relevant thread or issue details using commands like gh issue view ... --repo AgentWrapper/agent-orchestrator --json body,comments or live state via ao status/ao session ls.
- Minimum viable report gate: Ensure what happened, where, when, and (two of) OS/shell, AO version, reproducibility, reproduction steps are present; otherwise ask for more info.
- Local diagnostics (if bug on same machine): Run environment, daemon health, session details, and SQLite state using /tmp/ao and related commands; inspect daemon port 3001 and repository state at ~/.ao/data/ao.db.
- Investigate: Trace the actual code path across CLI, Daemon HTTP API, Sessions/Lifecycle, Runtime adapters, and Agent hooks; fetch commit history against upstream/main for context.
- Search for Duplicates: Use gh issue list and gh pr list with --repo AgentWrapper/agent-orchestrator and various --search terms to identify duplicates.
- If duplicate found: Comment on existing issue with a structured report including reporter, URL, date, and commit hash.
- If no duplicate: File a new issue with a detailed body including reproduction, root cause, fix plan, and impact; upload screenshots following provided commands; label and prioritize using existing labels, and include confidence and priority in the body.
- Push a fix PR: Create a local fix branch from upstream/main, build and test, push, and open a PR with a structured body including summary, test steps, and fixes references. If non-trivial, spawn a worker session to handle work in a separate worktree.
When to use it
Triggered when a user reports a bug or asks to triage/file an issue for a reported problem. It follows a step-by-step workflow to ensure issues are created with full context and linked to the correct upstream repository.
What it can touch
Commands and workflows explicitly mentioned include: git fetch upstream, git log --oneline upstream/main -5, ao status, ao version, /tmp/ao, gh issue list, gh issue comment, gh issue create, gh label list, gh pr create, and various git operations. It references path and binary usage: /tmp/ao as the verified rewrite binary, and the daemon port 3001. It uses repository AgentWrapper/agent-orchestrator.
Caveats
Requires using the verified rewrite binary and correct port (port: 3001) before trusting outputs. It emphasizes not filing on the origin fork and ensures triage targets the upstream repo. It includes conditional guidance for Windows/macOS/Linux platform considerations and cautions about duplicate detection and labeling limitations.
# Bug Triage Skill Triage bugs into well-structured GitHub issues on the upstream **`AgentWrapper/agent-orchestrator`** repo (issues are enabled there; the `origin` fork is not the issue tracker). > **Agent Orchestrator (AO) is Go + Electron.** The backend is a Go daemon > (`backend/`) exposing a loopback HTTP API on `127.0.0.1:3001`; the frontend is an > Electron + React supervisor (`frontend/`). There is **no** pm2/tmux-per-session > Node runtime here: the daemon owns lifecycle and terminals run under the **tmux** > runtime adapter (ConPTY on Windows). Triage against _this_ Go rewrite, not the old > TypeScript agent-orchestrator implementation. ## ⚠️ Which `ao` are you running? **A bare `ao` on your PATH may resolve to a different AO install** (for example an old npm build at `~/.nvm/.../bin/ao` that talks to port **:3000**). Triaging with the wrong binary produces bugs that don't exist in this rewrite (and misses ones that do). Before any diagnostics: ```bash which -a ao # see every ao on PATH; expect surprises ao status 2>/dev/null # if this shows port 3000, it is NOT this rewrite ``` Use a rewrite binary explicitly: ```bash # Option A: build from this repo (preferred during tr
- ⚠️ Which ao are you running?
- 1. Pre-flight
- 2. Gather Context
- 2a. Extract the report
- 2b. Minimum viable report gate
- 2c. Local diagnostics (if bug is on same machine)
- 3. Investigate
- 3a. Trace the code path
- 3b. Cross-platform check
- 3c. Stop-and-ask triggers
- 4. Search for Duplicates
- Duplicate found → comment on existing issue
- No duplicate → file new issue (next section)
- 5. File New Issue
which -a ao # see every ao on PATH; expect surprises ao status 2>/dev/null # if this shows port 3000, it is NOT this rewrite Option A: build from this repo (preferred during triage) cd backend && go build -o /tmp/ao ./cmd/ao Option B: the packaged app's bundled daemon Environment which -a ao # confirm no rogue ao shadows the build cat ~/.ao/running.json # PID + port handshake (expect port 3001) Daemon health lsof -i :3001 # who's bound to the daemon port
What does the bug-triage skill do?
Triage bugs reported in chat/issues, search for duplicates, file or update GitHub issues with full context, and push fix PRs.
How do I install it?
Run `npx skills add Untrivial-ai/agent-orchestrator --skill bug-triage --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From Untrivial-ai/agent-orchestrator, a repository with 8,786 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.