The Agent Skills Directory

[SAFE]: The skill is entirely declarative, providing a set of rules (FMT, BODY, FTR, REV, HYG) for the agent to follow when generating text. It does not perform any file system operations, network requests, or command executions.
[EXTERNAL_DOWNLOADS]: The skill references well-known and trusted external documentation (conventionalcommits.org and git-scm.com) for user education. These references are static URLs and do not involve runtime script execution.
[DATA_EXFILTRATION]: There is no evidence of sensitive data access or exfiltration. The skill operates solely on user-provided descriptions of code changes to produce commit messages.
[PROMPT_INJECTION]: The skill instructions are focused on formatting and do not contain any patterns intended to bypass safety filters or override system instructions.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns or dynamic script generation were detected. The skill does not use eval, subprocess, or network tools.

git-commits