Deployment Checklist

Generate a customized pre-deployment checklist based on what's being deployed and the risk level.

Gather Context

Before generating the checklist, determine:

1. What's changing? — New feature, bug fix, dependency update, infrastructure change, database migration, config change.
2. Risk level — Low (copy change, feature flag), Medium (new endpoint, schema change), High (auth changes, payment flow, data migration).
3. Deploy process — CI/CD auto-deploy, manual deploy, canary, blue-green, feature flag rollout.

If the user doesn't specify, ask. A config change and a database migration need very different checklists.

Checklist Template

Adapt this based on context — remove sections that don't apply, add specifics for the actual change.

Pre-Deploy

• Code reviewed and approved — PR has required approvals
• Tests passing — Unit, integration, and E2E tests green in CI
• No unrelated changes — PR is scoped to the stated purpose
• Environment variables — Any new env vars are set in staging and production
• Database migrations — Tested, reversible, and backward-compatible with current code
• Feature flags — New functionality behind a flag if applicable
• Documentation updated — API docs, runbooks, README if user-facing changes

Deploy

• Deploy to staging first — Verify in a non-production environment
• Smoke test staging — Hit critical paths manually or with automated checks
• Deploy to production — Follow the standard deploy process
• Canary period — If available, monitor canary before full rollout

Post-Deploy

• Smoke test production — Verify the change works in prod
• Monitor metrics — Error rates, latency, CPU/memory for 15-30 minutes
• Monitor logs — Watch for new exceptions or warnings
• Verify alerts — Confirm monitoring and alerting are active
• Communicate — Notify the team in the deploy channel

Rollback Plan

• Rollback command ready — Know exactly how to revert
• Rollback tested — Verified the rollback procedure works
• Rollback criteria — Define what triggers a rollback (error rate > X%, latency > Yms)
• Data rollback — If migrations ran, know how to reverse them

Risk-Specific Additions

Database migrations: Add backward-compatibility verification, migration dry-run, backup confirmation.

Auth/security changes: Add penetration test or security review, session handling verification, permission matrix check.

Payment/billing: Add sandbox transaction test, idempotency verification, reconciliation check.

Infrastructure: Add capacity planning review, DNS propagation check, certificate verification.

Output

Present as a clean, copy-pasteable checklist with the irrelevant sections removed and risk-specific items added. Include a one-line rollback command if the user shares their deploy tooling.