commit
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses local shell commands to interact with the git repository, including
git status,git diff,git log, andgit commitas defined in SKILL.md. - [SAFE]: A mandatory human-in-the-loop control is implemented, requiring the agent to display the drafted commit message and obtain user approval before proceeding with any write operations.
- [PROMPT_INJECTION]: The skill processes repository content that could potentially contain malicious instructions intended to influence the agent's behavior. 1. Ingestion points: Untrusted data is read from the output of
git diff --stagedandgit log. 2. Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands in the ingested data. 3. Capability inventory: The skill has the capability to execute repository modifications viagit commit. 4. Sanitization: No specific sanitization or filtering of the git output is described. Note that the required user review and approval of all generated messages serves as the primary security control.
Audit Metadata