The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a shell script find-polluter.sh designed to automate the process of finding which test file creates unwanted state or files. It uses npm test to execute local test files sequentially.
[COMMAND_EXECUTION]: The SKILL.md and root-cause-tracing.md files provide examples of diagnostic shell commands, such as env, security list-keychains, and codesign. These are presented as educational examples for identifying environment discrepancies in CI/CD or build pipelines.
[SAFE]: The skill includes several 'Pressure Test' markdown files. These are designed as adversarial scenarios to ensure the agent adheres to the systematic debugging process even under simulated time or social pressure. These function as robustness evaluations for the skill's logic.

systematic-debugging