The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documentation and the find-polluter.sh script utilize standard development commands such as npm test, find, codesign, and security list-keychains. These are used appropriately within the context of debugging build pipelines, test failures, and code signing issues on developer machines.- [DATA_EXFILTRATION]: Examples in SKILL.md demonstrate how to log environment variables and keychain identities to the local console. This is intended for diagnostic visibility during debugging and does not involve transmission of sensitive data to external servers.- [INDIRECT_PROMPT_INJECTION]: The find-polluter.sh script processes file paths retrieved from the local filesystem. While it handles external input (file names), it utilizes shell quoting to mitigate risks and is designed for execution by a developer in a local environment, posing no significant injection risk to the agent itself.

systematic-debugging