The Agent Skills Directory

[SAFE]: The skill provides a detailed debugging methodology across multiple phases (Build Loop, Minimise, Hypothesise, Instrument, Fix, Cleanup). It focuses on standard software engineering practices such as unit testing, performance profiling, and regression testing.
[SAFE]: The included bash script scripts/hitl-loop.template.sh is a benign utility for human-in-the-loop interaction. It uses standard read and printf commands to gather observations from a user and does not contain any suspicious network operations, credential harvesting, or unauthorized file access.
[SAFE]: The skill explicitly instructs the agent to redact secrets and auth headers from any output or captured artifacts, promoting good security hygiene.
[SAFE]: The YAML configuration in agents/openai.yaml is a standard interface definition with no dangerous tool permissions or configuration overrides.

diagnosing-bugs