| name | debug-prod |
| description | Debug production issues using cloud logs, CLI tools, error tracking, and database access. Use when debugging production errors, investigating service or job failures, or diagnosing masked errors in production. |
Debug Production
Help debug a production issue using cloud logs, error tracking, and database access.
Tools
Cloud CLI
Use your cloud provider's CLI for log queries and service inspection. See context.md for project-specific CLI
configuration, project IDs, and auth instructions.
Error Tracking
An error tracking service (Sentry, etc.) may be available as an MCP server. See context.md for usage guidance.
Database Access
First, check which Postgres MCP servers are available. If a production Postgres MCP is present and accessible,
query it directly (read-only). If only a dev server is present, do NOT query it for production debugging — ask the
user to run a specific SQL query on production.
Workflow
- Gather context — Ask for symptom, error message, affected users/endpoints, time window. Check recent deploys.
- Identify the source — Determine service vs background job. If unclear, use cross-resource log search.
- Pull logs — Use appropriate query template. Start with ±2 min window. Search for application logs.
- Correlate — Cross-reference with recent code changes (
git log), deployment history, service health.
- Hypothesize and verify — Form root cause hypothesis, find supporting evidence before proposing a fix. Since
production errors are often masked, trace the code path.
- Propose a fix — Present minimal, targeted fix with clear reasoning. Flag risks or rollback considerations.
Project Context
Read context.md and apply it as additional project-specific constraints layered on top of this
workflow. If it does not exist, skip this section.
