Capgo Incident Response

Use this skill when a live update causes regressions, startup failures, or elevated crash rates.

Triage Sequence

1. Confirm incident scope (platforms, channel, app version, error profile).
2. Fetch currently linked bundle for impacted channel.
3. Identify previous known-good bundle.
4. Decide rollback now vs partial containment.

Immediate Command Toolkit

npx @capgo/cli@latest channel currentBundle <channel> <appId>
npx @capgo/cli@latest bundle list <appId>
npx @capgo/cli@latest channel set <channel> <appId> --bundle <knownGoodBundle>
npx @capgo/cli@latest app debug <appId>

Containment Patterns

• Roll back production channel to last good bundle.
• Restrict channel targeting if issue is platform-specific.
• Pause promotion pipeline until compatibility gates pass again.

Post-Incident Actions

1. Record timeline with first bad bundle and rollback timestamp.
2. Run compatibility checker on the failed bundle.
3. Add missing guardrails/tests that would have caught the issue.
4. Prepare fix-forward bundle and redeploy through staged channels.

Output Template

• Incident summary
• Impact
• Actions taken
• Current state
• Root cause hypothesis
• Preventive controls