fix(goap): orphaned skill — bug_triage routed to non-existent auto-triage-sweep agent

[mabry1985]

Summary

GOAP dispatches bug_triage work items to an agent called auto-triage-sweep, which is not registered in the agent health registry. This creates an orphaned skill (orphanedSkillCount: 1) and a cascading failure loop.

Impact

• 100% failure rate on auto-triage-sweep (4/4 failed)
• 11 items stuck in error state, WIP blown to 24 (limit: 2)
• 12 duplicate incidents (INC-003 through INC-016) filed by the GOAP feedback loop, all describing the same stuck agent
• GOAP keeps re-dispatching, creating a tight failure→alert→dispatch→failure loop

Root Cause

The auto-triage-sweep agent is referenced in GOAP action config but does not appear in the live agent registry. It was either:

• Never deployed
• Deregistered / crashed without cleanup
• A config typo

Meanwhile, Quinn (quinn) is online, reachable (63ms latency), and advertises bug_triage as a skill.

Fix

1. Re-route the GOAP action for bug_triage from auto-triage-sweep → quinn (or whichever agent should own this)
2. Drain the 11 error-state items (retry or discard)
3. Consolidate/close INC-003 through INC-016 once routing is fixed
4. Add a guard in GOAP dispatch: if the target agent is not in the live registry, skip dispatch and fire an alert instead of looping

Labels

bug, goap, priority:high

[github-actions]

👋 Thank you for submitting this idea!

Your issue has been added to our ideas pipeline. Here's what happens next:

1. Triage (1-3 days): Our team will review and label your idea
2. Discussion (ongoing): We may ask clarifying questions or gather community feedback
3. Acceptance (varies): If accepted, we'll add the status: accepted label
4. Implementation (varies): Our AI agents will implement accepted ideas and link the PR here

We appreciate your contribution to making this project better! 🚀

[protoquinn]

Triage — Quinn (QA)

Verdict: CONFIRMED HIGH — recurring pattern, 12th documented instance

---

Validation

• Root cause confirmed: actions.yaml references auto-triage-sweep as the dispatch target for bug_triage. This agent is not present in the live fleet registry.
• Quinn (quinn) is online, reachable, and advertises bug_triage as a skill — correct re-route target.
• QA memory cross-reference: this exact failure mode is documented in bugs feat(server): add scheduler service for cron-based task execution #34, Add scheduled periodic sync task #37, Auto-sync before agent starts feature work #39, GitHub webhook to auto-close features on PR merge #41, [Epic] Scheduled Task Persistence #42, [Epic] Agent Resume Integration #45, Create Default PR Feedback Hook #50, Add /pr-review plugin command #55 (and now Add /pr-review plugin command #55 explicitly labels this the "11th instance"). This is the 12th.
• Duplicate issues filed same day: fix(goap): remove stale auto-triage-sweep routing — orphaned bug_triage skill causing feedback loop #3530, fix(goap): remove orphaned auto-triage-sweep skill registration — route bug_triage to Quinn #3531 describe the same root cause.

Impact confirmed

• 100% dispatch failure rate on auto-triage-sweep (4/4 failed)
• 11 items stuck in error state, WIP at 24 (limit: 2)
• 12 duplicate incidents (INC-003 through INC-016) from GOAP feedback loop
• Self-reinforcing failure→alert→dispatch→failure cycle active

Board bug filed: feature-1776669537453-z2uls3cvs (severity: high)

Recommended fix (in priority order)

1. actions.yaml: re-route bug_triage dispatch target from auto-triage-sweep → quinn
2. Drain the 11 error-state items (retry or discard)
3. Consolidate INC-003 through INC-016 — close all but the canonical incident
4. Add pre-dispatch registry guard in GOAP: if target agent is not in live registry, skip dispatch and fire a single alert instead of looping

Duplicates

Issues #3530 and #3531 describe the same root cause. Recommend closing both as duplicates of this issue.

Owner: Kai (backend) for the GOAP dispatch guard and actions.yaml re-route; Sam (agent flows) for the incident dedup logic.

[protoquinn]

Triage — Quinn (QA)

Confirmed. This is a verified, recurring root cause. QA memory shows 12+ prior instances of the same GOAP auto-triage-sweep orphaned routing pattern (bug patterns #37, #41, #42, #55, #57, #62, #70, #75 in QA store). The fix has not been applied to actions.yaml across any of those cycles — this is a regression, not a new bug.

Severity: HIGH (100% dispatch failure rate, 51 items stuck in error state, 14 duplicate incidents generated)

Root cause: actions.yaml routes bug_triage skill to executor target auto-triage-sweep, which is not registered in the live agent fleet. Every dispatch fails, GOAP retries, failure cascades.

Impact:

• 4/4 outcomes failed (0% success rate)
• 51 items in error state in the flow pipeline
• 14 duplicate incidents (INC-003 through INC-016) filed for the same root cause
• GOAP remediation loop cannot converge

Fix required (owner: Kai / Frank):

1. Remove or reroute bug_triage in actions.yaml — valid targets: quinn or protomaker (both advertise bug_triage)
2. Add pre-dispatch registry validation in the GOAP dispatcher — skip actions targeting agents not in the live registry, log a warning instead of retrying
3. Investigate protopen A2A card returning null (security skills orphaned) — tracked in fix(goap): remove stale auto-triage-sweep agent from action routing — orphaned skill causing 100% failure loop #3533 context
4. jon empty skills array tracked in fix(jon): A2A agent card advertises zero skills — causes orphaned skill in fleet health #3513

Duplicate: #3533 covers the same root cause with additional context — closing that as duplicate of this issue.

Filing board bug now.