post-deploy

name	post-deploy
description	Post-deployment smoke tests and notification workflow. Verifies critical paths are working and notifies the team of deployment status.

Post-Deploy Verification

Run smoke tests against a freshly deployed environment to verify critical user paths are working, then notify the team of the deployment status.

When to Use

Use this skill when:

• A deployment to staging or production just completed
• After a rollback to verify the previous version works
• As part of an automated deployment pipeline
• When users report issues after a deploy

Instructions

1. Identify the deployment target:

   • Ask for the environment URL if not provided (e.g., https://staging.example.com)
   • Determine the environment type: staging | production
   • Get the deployed version/commit hash if available

2. Run health checks:

   # Basic connectivity
   curl -sf {URL}/api/health | jq .
   
   # Response time baseline
   curl -o /dev/null -s -w '%{time_total}\n' {URL}
   
   # SSL certificate validity
   echo | openssl s_client -servername {domain} -connect {domain}:443 2>/dev/null | openssl x509 -noout -dates
   

   Expected results:

   • Health endpoint returns 200 with {"status": "ok"}
   • Response time under 500ms for initial page load
   • SSL certificate valid for at least 14 days

3. Test critical user paths:

   Authentication

   • Login page loads without errors
   • OAuth redirect URLs are correct for this environment
   • Session/token handling works (no auth loops)

   Core Features

   • Homepage renders with dynamic content (not stale cache)
   • API endpoints return expected data shape
   • Database queries execute (test a list endpoint)
   • File uploads work (if applicable)
   • WebSocket connections establish (if applicable)

   Error Handling

   • 404 page renders correctly for unknown routes
   • API returns proper error format for invalid requests
   • Rate limiting is active (if applicable)

4. Check infrastructure:

   • Verify error tracking is receiving events (Sentry, Bugsnag, etc.)
   • Check application logs for errors in the last 5 minutes
   • Verify metrics/monitoring is reporting (Datadog, New Relic, etc.)
   • Check CDN cache is properly invalidated for updated assets
   • Verify cron jobs / background workers are running

5. Generate deployment notification:

   For Slack / Discord

   🚀 Deployed to {environment}
   
   Version: {commit_hash_short}
   Branch: {branch}
   Deployed by: {deployer}
   Time: {timestamp}
   
   Smoke Tests: ✅ All passing (or ❌ X failures)
   Health: {response_time}ms
   
   Changes:
   - {commit_message_1}
   - {commit_message_2}
   
   Dashboard: {monitoring_url}
   

   For GitHub (Release / PR comment)

   Post a comment on the merged PR or create a release note with the smoke test results.

6. Output a post-deploy report:

   ## Post-Deploy Report: {environment}
   
   | Check          | Status | Details              |
   |----------------|--------|----------------------|
   | Health         | ✅     | 200 OK, 127ms        |
   | Auth           | ✅     | Login flow works     |
   | Core Features  | ✅     | 5/5 paths verified   |
   | Error Tracking | ✅     | Sentry connected     |
   | SSL            | ✅     | Valid for 83 days    |
   
   ### Deployment: HEALTHY ✅
   

7. If issues are detected:

   • Clearly flag which checks failed
   • Provide specific error messages and logs
   • Recommend rollback if critical paths are broken
   • Draft a rollback command: git revert {commit} && git push

Example Usage

Run post-deploy checks on https://staging.myapp.com

Verify production is healthy after the latest deploy

Generate a deploy notification for Slack