agentsclimarketplace

Launch checklist

Skill aroyburman-codes/pm-skills/skills/launch-checklist

PM workflow and product thinking skills for AI product managers. 17 structured frameworks for PRDs, metrics, strategy, writing, prioritization, and more.

Install
npx -y skills add aroyburman-codes/pm-skills --skill launch-checklist

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 15 stars15 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate launch readiness checklists for product releases. Covers engineering, QA, design, legal, marketing, support, and rollback planning. Adapts to launch size (feature flag, beta, GA) and AI/ML-specific requirements.

SKILL.md

4.1 KB, as published. Nobody here has run it

Launch Checklist Skill

Generate a comprehensive, customized launch readiness checklist for any product release.

When to Use

  • User is preparing to launch a feature or product
  • User needs to ensure nothing is missed before go-live
  • User says /launch-checklist followed by what they're launching
  • Any pre-launch review or go/no-go decision

Framework: Launch Readiness Checklist

Step 1: Scope the Launch

  • What's launching? Feature name and brief description
  • Launch type: Feature flag → Internal → Beta → GA
  • Target date: When is this going live?
  • Blast radius: How many users are affected?
  • Reversibility: Can we roll back easily?

Step 2: Generate Checklist by Function

Engineering

  • Code complete and merged to release branch
  • Unit tests passing (coverage threshold met)
  • Integration tests passing
  • Load/performance testing complete (meets latency/throughput targets)
  • Feature flag configured and tested (on/off/gradual rollout)
  • Database migrations tested and reversible
  • API versioning handled (no breaking changes for existing clients)
  • Monitoring and alerting configured
  • Logging sufficient for debugging
  • Rollback procedure documented and tested

QA

  • Test plan executed (happy path, edge cases, error states)
  • Cross-browser/cross-platform testing complete
  • Accessibility testing (screen reader, keyboard nav, contrast)
  • Regression testing on adjacent features
  • Security review complete (auth, input validation, data handling)

Design

  • Final designs match implementation
  • Empty states, loading states, error states all designed
  • Mobile/responsive behavior verified
  • Copy reviewed and finalized
  • Animations/transitions smooth and performant

Legal & Compliance

  • Privacy review complete (data collection, retention, deletion)
  • Terms of service updated if needed
  • GDPR/CCPA compliance verified
  • Licensing for any third-party components cleared

Marketing & Communications

  • Launch announcement drafted
  • Help center / documentation updated
  • In-app messaging configured (tooltips, onboarding)
  • Social media / blog post scheduled (if applicable)
  • Internal comms sent to stakeholders

Support

  • Support team briefed on new feature
  • FAQ and troubleshooting guide prepared
  • Escalation path defined for launch-related issues
  • Known limitations documented

AI/ML-Specific (if applicable)

  • Model eval results meet quality bar
  • Safety/red-team testing complete
  • Content policy and guardrails configured
  • Hallucination rate within acceptable threshold
  • Cost per query within budget
  • Fallback behavior for model errors/timeouts
  • Data pipeline monitoring active
  • Bias testing on key demographic segments

Step 3: Go/No-Go Criteria

Define clear criteria for each launch stage:

Internal Launch Go/No-Go:

  • All P0 bugs resolved
  • Core user flow works end-to-end
  • Monitoring active

Beta Launch Go/No-Go:

  • All P0 and P1 bugs resolved
  • Performance meets targets
  • Support team briefed
  • Rollback tested

GA Launch Go/No-Go:

  • All P0-P2 bugs resolved
  • Load testing passed at 2x expected traffic
  • Documentation complete
  • Legal review signed off
  • Monitoring and alerting verified

Step 4: Rollback Plan

  • Trigger criteria: What signals would trigger a rollback?
  • Rollback steps: Exact steps to revert (feature flag off, DB rollback, etc.)
  • Communication: Who to notify and how
  • Timeline: Maximum time from decision to rollback complete

Output Format

Generate as a markdown checklist grouped by function. Mark items as applicable/not applicable based on the specific launch. Include a summary section with launch date, owner, and go/no-go criteria.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.