washingtonguilhardes/review

REVIEW Task

Persona: Execute this task as the @qa subagent. Load the persona characteristics from .rulesync/subagents/qa.md before proceeding.

Objective

QA review of PR/branch. Execute checks, apply checklists, identify issues, produce report. Save to docs/qa/reports/{yyyy-mm-dd}-{branch-or-pr-slug}.md.

Instructions

1. Initiate: - Greet as Quinn - Ask: "What to review? (PR URL, branch name, or describe changes)" - If PR URL + GitHub MCP: fetch PR details - If branch: git diff to see changed files - Ask the user to start the dev server if it is not already running.

2. Automated checks: - pnpm lint && pnpm typecheck && pnpm test && pnpm test:e2e (skip E2E if not feasible, note in report) - Record pass/fail and errors

3. Code quality: - Check: console.log/debug statements, any types, error handling, sensitive data in logs, JSDoc on exports, import order, naming - Search for obviated code to remove - Identify opportunities to consolidate similar code

4. Quality gate checklist: - Primary: Code Quality, Testing, Documentation, Performance - Secondary (only if explicitly prioritized by project rules): Security, Accessibility

5. Additional context: 1. Review related briefs/decisions/specs in docs/ and ensure the functionality is fully-implemented, unfinished work should be categorized as P0 2. Manual testing? Describe scenarios 3. UI changes? 4. Screenshots/videos? 5. Known issues/limitations?

6. Categorize issues: - P0 - Critical: Blocks merge (failing tests, security vulns, broken functionality) - P1 - High: Fix before merge (missing error handling, a11y gaps) - P2 - Medium: Address soon (minor code quality, missing docs) - P3 - Low: Nice to have (code style, optimizations)

7. Decision: - PASS: All critical checks pass, no blockers - PASS WITH RECOMMENDATIONS: Can merge, non-blocking issues - BLOCK: Critical issues must fix

8. Generate report: - Read template: .rulesync/templates/qa-report-template.md - Follow template structure exactly - Fill sections: findings, automated check status, issues with severity/fixes, decision with rationale - Save to docs/qa/reports/{yyyy-mm-dd}-{branch-or-pr-slug}.md

9. Summary: - Decision summary - Critical findings - Required actions if blocked - Report link - If blocked: "PR blocked. Address critical issues." - If passed: "PR passes QA review, ready to merge."

10. Rule improvement analysis: - Identify gaps: patterns not covered, repeated issues, new best practices, patterns needing explanation - Generate suggestions: Target rule domain (e.g., "unit testing", "code quality", "architecture"), section, type (Addition/Update/New Pattern), reasoning, proposed content, examples - Ask: "Identified {N} potential rule improvements. Review and apply?" - If yes: Show each, ask approve/reject, apply approved updates, track updates - If rules updated: Confirm updates applied - Summary: Total suggestions, applied, skipped, updated files, impact, next steps

TODO Composition

Create todos at task start:

1. review-initiate - "Greet Quinn and identify what to review"
2. review-automated-checks - "Run automated checks (lint, typecheck, tests)"
3. review-code-quality - "Review code quality (console.log, any types, error handling, JSDoc, etc.)"
4. review-quality-gate - "Apply quality gate checklist (code, testing, docs, security, performance, accessibility)"
5. review-additional-context - "Gather additional context (manual testing, UI changes, screenshots, known issues)"
6. review-categorize-issues - "Categorize issues by severity (P0/P1/P2/P3)"
7. review-decision - "Make decision (PASS/PASS WITH RECOMMENDATIONS/BLOCK)"
8. review-generate-report - "Generate QA report from template"
9. review-summary - "Generate review summary"
10. review-rule-improvements - "Analyze and apply rule improvements"

Update status: Mark in_progress when starting each, completed when done.