The Agent Skills Directory

[SAFE]: The skill provides a structured debugging methodology and includes a utility script (find-polluter.sh) for test bisection. The script uses standard shell commands such as find and npm test to isolate problematic tests, which is consistent with its stated purpose.
[SAFE]: The documentation contains illustrative examples of diagnostic commands, including environment variable inspection and keychain queries (security list-keychains). these are used in context as examples for debugging CI/CD and build failures and do not constitute a data exfiltration risk.
[SAFE]: The skill includes several pressure-test scenarios (test-pressure-1.md, etc.) designed to train the agent to resist shortcuts and follow systematic procedures under constraints. These scenarios improve the reliability and safety of the agent's problem-solving behavior.
[SAFE]: No obfuscation, remote code execution patterns, or unauthorized network operations were detected. All provided scripts and instructions are transparent and aligned with technical troubleshooting best practices.

systematic-debugging