Debugging Production Errors Skill Reference

Overview

debugging-production-errors helps a coding agent investigate production incidents with the project’s live log connection. Use it when a failure, spike, or customer report needs evidence from recent logs before the next fix lands.

Skill contract

  • It reads logs only and never changes log data.
  • It works only after the project has a configured log connection.
  • If no connection exists, your agent points to the project’s Logs settings UI and ends the investigation.
  • It keeps the search focused on the relevant project, service, time range, and failure pattern.
  • It can surface common failure modes such as request timeouts, rate limits, truncated results, connection errors, and noisy unstructured lines.

Workflow

  1. Confirm the configured connection. Your agent checks that the project can reach its log source before any deeper investigation.
  2. Scope the incident. Your agent narrows the search to the affected service, time window, and symptom.
  3. Inspect the failure. Your agent looks for the strongest error pattern, the first occurrence, and any repeated signal.
  4. Gather evidence. Your agent collects the relevant log lines, stack traces, and surrounding context.
  5. Map the findings back to the source code or the relevant fix. Your agent ties the evidence to the changed files, the deploy, or the code path that likely caused the issue.

See also