The Agent Skills Directory

[SAFE]: The skill strictly defines a debugging workflow, including instructions for redacting secrets (e.g., <REDACTED>) before displaying logs or artifacts. It emphasizes building feedback loops using standard tools like unit tests, cURL, and Playwright. The provided shell script is a simple template for human-in-the-loop interaction, containing no automated execution of dangerous or external commands. All logic focuses on internal code diagnosis and follows industry-standard security best practices for handling sensitive data during troubleshooting.

diagnosing-bugs