Handoff checklist before a consultant joins your incident
External help works when internal teams front-load context. This checklist is what we send before every debugging block at Node Grove.
People
- One decision-maker who can approve restarts or rollbacks
- One engineer who deployed the last change
- Optional DBA or infra contact on standby, not in the main thread unless needed
Access
- Read-only repo access to the deployed tag
- Log search for ±2 hours around first alert (redacted)
- Staging environment that reproduces the bug, or explicit note that it does not
Artifacts
- Timeline in UTC: deploys, config changes, traffic anomalies
- Sample failing request IDs and expected vs actual response
- Known blind spots ("we do not trace worker processes")
Completing this list usually saves the first paid hour for diagnosis instead of access chasing. See our full engagement process.