pr-review-guardrails

Stress-test the primary review with an additional independent reviewer that generates its own findings, compares reviewer conclusions, and identifies issues the primary reviewer may have missed. Use when performing a second opinion or double-check review on a pull request, for medium or high risk PRs, when authoring was heavily AI-assisted, when primary reviewer confidence is low, when findings conflict, or when you need to verify findings with a cross-model or same-model challenger. Supports same-model and cross-model configurations for fair comparison.

Generates a structured, human-readable reviewer packet summarising what changed in a pull request, why it matters, what was verified, and where human attention is most needed. Use when the user asks for a PR review summary, a code review packet, a human-readable change report, or wants to hand off review findings to a human reviewer. Produces a scannable document: quick approvals (low-risk PRs) can be assessed in under 30 seconds; detailed reviews (high-risk PRs) in under 2 minutes. Outputs a formatted markdown packet with risk rating, verification status, ranked findings, unresolved questions, and a recommended review focus — making human review faster without replacing human judgment.

Build a compact, trustworthy evidence pack before deeper PR review starts. Use this skill when a pull request needs review — it is always the first step. Triggered by requests to review code, check a PR, review my changes, review a merge request, or any similar code review or pull request review request. Collects PR context, runs deterministic verifiers, classifies risk, maps hotspots, and checks for missing artifacts. Produces the evidence pack that all downstream review skills consume.

Evaluates which code review comments (review tiles) actually produced changes after a pull request is merged or closed, by passively collecting outcome data from the GitHub API and git history — zero developer friction. Use when analyzing post-merge pull request outcomes, assessing code review effectiveness, measuring review feedback impact, or answering questions like "how did PR #6 go?", "which review comments were accepted?", or "did any escaped defects appear after this pull request merged?" Produces a structured per-finding outcome record (accepted / rejected / ignored / superseded), merge time delta, escaped defect count, and AI authorship correlation for each PR.

Turn many candidate findings from reviewers and verifiers into a small, decision-useful set. Deduplicates, ranks, and suppresses weak findings to consolidate review results into a prioritized, actionable list with severity ratings and merged confidence scores. Use when you need to merge findings, consolidate feedback, prioritize issues, or summarize review output after review passes are complete and before human handoff. Trigger phrases: "consolidate review results", "merge findings", "deduplicate feedback", "prioritize issues from review", "summarize reviewer output". The evidence threshold is the filter — not an arbitrary cap.

Provide an independent critique of a pull request (PR) using a clean reviewer context, identifying bugs, security issues, code quality problems, API misuse, and missing test coverage. Use when performing a code review or pull request review after an evidence pack has been built, for green or yellow risk lane PRs, or as part of a full pipeline for red risk lane PRs. Produces candidate findings (covering correctness, security, and architectural concerns) for downstream synthesis — not final verdicts. Operates as a critic, not a co-author. Common triggers: "review this PR", "code review feedback", "fresh review", "independent review".