conventional-commit
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No patterns detected that attempt to bypass AI constraints or override system instructions.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or unauthorized network requests were found.
- [Obfuscation] (SAFE): The content is standard markdown and does not utilize encoding, zero-width characters, or homoglyphs.
- [Unverifiable Dependencies & Remote Code Execution] (SAFE): No package installations or remote scripts are referenced or executed.
- [Command Execution] (SAFE): The included git commands (e.g.,
git branch --show-current) are standard read-only or local-only development commands relevant to the skill's purpose. - [Indirect Prompt Injection] (SAFE): While the skill assists in generating commit messages from user input, it lacks dangerous capabilities (like file writes or network ops) that could be exploited via injection.
Audit Metadata