conventional-commits
Pass
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use standard Git tools (git status, git diff) to read the local repository state. These commands are necessary for the skill's purpose and do not pose a security risk.
- [PROMPT_INJECTION]: Instructions are focused on formatting and linguistic requirements for commit messages. No patterns of jailbreaking, safety bypass, or instruction overrides were detected.
- [DATA_EXFILTRATION]: The skill does not contain any network-related commands or logic to transmit data to external servers.
- [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or sensitive environment variable access were found in the provided files.
Audit Metadata