The Agent Skills Directory

[SAFE]: The skill implements proactive security guardrails by checking for sensitive files such as .env, .pem, and files containing 'secret', 'token', or 'password' in the repository changes, providing a warning to the user to prevent accidental data exposure.
[COMMAND_EXECUTION]: The skill uses standard read-only git commands (git status, git diff, git log) to analyze project context. It includes explicit instructions to never execute modifying commands like git add, git commit, or git push, ensuring the agent remains in a advisory role.
[DATA_EXPOSURE]: While the skill reads local repository data to generate suggestions, it does not perform any network operations or external data transfers, keeping all analyzed information within the local environment.

commit-msg