Consolidate overlapping PR skills-review agents: Matt Pocock Skills Reviewer + Impeccable Skills Reviewer

[github-actions]

Warning

Threat Detection Engine Failure — The analysis engine could not complete. This is a tooling failure, not a security finding.

What happened

The threat detection engine failed to produce results.

Review the workflow run logs for details.

Overlap Detected

Matt Pocock Skills Reviewer (.github/workflows/mattpocock-skills-reviewer.md) and Impeccable Skills Reviewer (.github/workflows/impeccable-skills-reviewer.md) both:

• trigger on pull_request: ready_for_review with the same doc/scratchpad path-ignore list
• pre-fetch the same PR data (pr-meta.json, pr-diff.patch, pr-review-comments.json) via shared/pr-diff-data-fetch.md
• select 1-2 relevant skills, post up to 10 inline create-pull-request-review-comments, submit one submit-pull-request-review, and optionally one add-comment
• share nearly identical scope rules (changed-lines-only, security-first priority, skip generated files, progressive disclosure)

Both are currently healthy at runtime (100% success in the latest metrics window: Matt Pocock 19/19, Impeccable 20/20), so this is not a reliability problem — it is duplicate reviewer coverage on every qualifying PR, which increases noise and token cost.

Recommendation

1. Decide whether the two skill sets (mattpocock/skills vs Impeccable skills) are meant to be complementary or redundant.
2. If complementary: split responsibilities explicitly in each prompt (e.g., one covers TDD/bug-diagnosis, the other covers a distinct skill domain) so they do not both comment on the same PR.
3. If redundant: merge into a single workflow that loads both skill sources and dedupes findings, or deprecate one.

Evidence

• Workflow definitions: .github/workflows/mattpocock-skills-reviewer.md, .github/workflows/impeccable-skills-reviewer.md
• Metrics: /tmp/gh-aw/repo-memory/default/metrics/latest.json (timestamp 2026-08-06T03:44:10Z)

Filed by Agent Performance Analyzer as part of the weekly performance report.

Generated by ⚡ Agent Performance Analyzer - Meta-Orchestrator · auto · 36 AIC · ⊞ 16.9K · ◷

• expires on Aug 8, 2026, 5:18 AM UTC-08:00

[github-actions]

🍪 Issue Monster selected this for Copilot

I've identified this issue as a good candidate for automated resolution and requested assignment to the Copilot coding agent.

If assignment succeeds, the Copilot coding agent will analyze the issue and create a pull request with the fix.

Om nom nom! 🍪

🍪 Om nom nom by Issue Monster · gpt54 · 5.83 AIC · ⊞ 10.1K · ◷