[pr-review] PR #11137 Cycle 1

Status: Approved

Executive Summary

PR #11137 is the right-sized fix for #11136: it names the invoked skills literally in the two highest-traffic peer-routing surfaces, without adding schema, lifecycle ceremony, or broad substrate text. I found no required actions.

The PR's empirical anchor is directionally valid: PR #11127 had substantive issue-comment text but formal review bodies that were empty or placeholder-level, which supports the "semantic routing was too weak" premise.

Required Actions

None.

Verification Performed

• gh pr view 11137 --json state,title,headRefName,headRefOid,baseRefName,reviewDecision,latestReviews,comments,body,url
• gh pr checks 11137
• git cat-file -e b9d15587a2f0a51c5dcf963b9ca42f948122686d^{commit}
• git diff --check origin/dev...b9d15587a2f0a51c5dcf963b9ca42f948122686d
• gh pr diff 11137 --patch
• gh issue view 11136 --json title,body,labels,state,url
• gh pr view 11127 --json comments,latestReviews,reviewDecision
• gh api repos/neomjs/neo/pulls/11127/reviews
• Targeted grep at PR head for old/new routing phrases.

CI / Security Audit

Pass.

Current gh pr checks 11137:

• Analyze (javascript): pass
• CodeQL: pass
• integration-unified: pass
• unit: pass

No local test rerun: this is a docs/workflow text-only change, and the active Memory Core recovery incident is still under read-only / no-test discipline. Per the review skill's docs-only rule, exact diff review plus git diff --check and green hosted CI is sufficient here.

Acceptance Criteria Audit

Pass.

Against #11136:

• Pull request workflow now literally requests /pr-review in the A2A body.
• Ideation sandbox workflow now literally requests /peer-role for design-review context and /ideation-sandbox for co-authoring divergence.
• Diff is +2/-2, well under the under-10-line scope guard.
• No new structured A2A fields, no schema, no daemon behavior, no MCP changes.
• PR body links the #11127 empirical anchor and keeps post-merge validation separate from the static review surface.

Architecture / Scope Audit

Pass.

This stays at the right abstraction level: the problem is skill auto-fire reliability at the prompt/message boundary, not A2A transport shape. Naming the skill literally in existing human-readable body text is the smallest substrate mutation that changes agent behavior without burdening every message with additional fields.

I considered whether this should also update every reviewer-request phrase in the repo. I do not think that should block this PR. The ticket is explicitly scoped to the two hot paths, and widening it would weaken the net-reduction discipline this patch is trying to preserve.

Source-of-Authority Audit

Pass.

The PR body cites #11136 as the implementation source of authority and #11127 as empirical evidence, not as an operator quote substitute. I verified both surfaces before accepting the framing.

Evidence Audit

Pass.

This PR has static docs/workflow acceptance criteria. The evidence ladder does not require runtime substrate proof for this patch. The PR body's post-merge validation target is useful and correctly framed as a future empirical check, not as a pre-merge claim.

Cross-Skill Integration Audit

Pass.

The modified producer surfaces now align with the skill routing table already present in AGENTS.md: PR review requests name /pr-review, and design-review discussion requests name /peer-role. No downstream consumer contract needed updating because the change is literal text in existing A2A request bodies.

Strategic-Fit Decision

Approve.

This is a compact map correction, not a new atlas page. It directly reduces the chance of another PR #11127-style rubber-stamp by making the expected skill invocation machine-obvious to the receiving agent.

Metrics

• ARCHITECTURAL_ALIGNMENT: 95
  • Correct layer and no schema inflation. Small residual is that broader phrase cleanup is intentionally deferred.
• CONTENT_COMPLETENESS: 95
  • Covers #11136 ACs cleanly. The PR #11127 anchor is directionally accurate; the formal review bodies were one empty and one placeholder-level rather than both strictly empty.
• EXECUTION_QUALITY: 95
  • Exact-head diff reviewed, whitespace check passed, hosted CI green. No local tests by design for docs-only plus active incident discipline.
• PRODUCTIVITY_IMPACT: 100
  • All explicit ACs are satisfied with a two-line diff and no additional operational load.
• IMPACT: 65
  • Meaningful peer-routing reliability improvement, but deliberately narrow.
• COMPLEXITY: 15
  • Minimal workflow text change.
• EFFORT_PROFILE: Quick Win

Risk / Follow-Up

Non-blocking: after merge, the real proof is whether the next peer-routing request actually auto-fires the intended skill. The PR body already names that validation target, so I would leave this PR as-is and use the next miss, if any, as the follow-up signal.

Merge Authority

Approved for human merge consideration only. Per repository invariant, I am not merging this PR.

Review Metadata

• PR: #11137
• Head: b9d15587a2f0a51c5dcf963b9ca42f948122686d
• Base: dev
• Cycle: 1