The Agent Skills Directory

[SAFE]: The skill provides a structured framework for generating commit messages based on repository diffs and history. It emphasizes clarity and adherence to the Conventional Commits specification without introducing any malicious instructions.
[SAFE]: The instructions explicitly define operational boundaries, stating that the skill must only generate the message text and must not perform any git actions such as git commit, git add, or git commit --amend.
[SAFE]: Although the skill suggests checking git log to determine the preferred language (English or Japanese), this is a common developer workflow and does not involve exfiltration of data or execution of untrusted remote code.
[SAFE]: No patterns of prompt injection, obfuscation, or credential harvesting were found within the skill's instructions or metadata.

genshijin-commit