The Agent Skills Directory

[SAFE]: The skill is designed for a specific, non-destructive task and does not exhibit any malicious patterns or unauthorized behaviors.
[COMMAND_EXECUTION]: The skill uses git diff and git diff --cached to read the state of the repository. These commands are executed via bash but are restricted to read-only operations and are explicitly defined in the skill's allowed tools.
[PROMPT_INJECTION]: While the skill ingests untrusted data from git diffs (Indirect Prompt Injection surface), the potential impact is mitigated by the skill's lack of dangerous capabilities (like network or write access) and the strict formatting requirements for its output.

commit