The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses subprocess.run to interact with Git and execute its own validation scripts. All commands use static arguments or controlled inputs, posing no risk of command injection.
[INDIRECT_PROMPT_INJECTION]: The skill processes git diffs, which could contain untrusted content. The implementation manages this risk by focusing on metadata and structure for batching, rather than executing content from the diff, and by including safety checks for risky patterns.
[SAFE]: The skill follows secure coding practices, implementing safety gates to detect and prevent the accidental exposure of sensitive credentials and local artifacts in git history.

conventional-commit-batcher