Set up an AI-orchestrated QA and ship checklist

Description

You're about to ship a feature and need a comprehensive pre-launch checklist that goes beyond manual testing. With AI-assisted QA becoming the norm, define what your review-test-ship pipeline should actually cover — from code review to post-deploy canary monitoring.

You are a senior PM setting up a comprehensive AI-assisted QA and ship pipeline for a product team. The goal is to define a repeatable checklist that leverages AI tools at every stage, ensuring nothing ships without proper verification.

## Feature Context
- Feature name: {{feature_name}}
- Product: {{product_name}}
- Scope: {{scope_summary}}
- Target ship date: {{ship_date}}

## Phase 1: Pre-Review Checklist
Before requesting code review:
- [ ] All acceptance criteria from the PRD are implemented
- [ ] Unit test coverage meets threshold (specify: {{coverage_target}})
- [ ] No hardcoded secrets, credentials, or PII in the diff
- [ ] Database migrations are reversible
- [ ] API changes are backward-compatible or versioned

## Phase 2: Code Review Gates
Define what an AI-assisted code review should catch:
- [ ] SQL injection, XSS, and OWASP Top 10 vulnerabilities
- [ ] Conditional side effects (state changes in if-branches that could silently fail)
- [ ] LLM trust boundary violations (if applicable — user input flowing into prompts)
- [ ] Performance regressions (N+1 queries, unbounded loops, missing pagination)
- [ ] Completeness: are there TODO comments or unfinished paths?

## Phase 3: QA Testing Protocol
Define the test matrix:
1. **Happy path** — Core user flow works end-to-end
2. **Error states** — What happens when the API is down, input is invalid, permissions are wrong?
3. **Edge cases** — Empty states, maximum input lengths, concurrent users
4. **Responsive** — Test at mobile (375px), tablet (768px), desktop (1440px)
5. **Accessibility** — Keyboard navigation, screen reader, color contrast
6. **Cross-browser** — Chrome, Safari, Firefox minimum

## Phase 4: Ship Checklist
- [ ] Feature flag configured (if gradual rollout)
- [ ] Monitoring/alerting set up for key metrics
- [ ] Rollback plan documented (how to revert in

Customize This Prompt