commit-messages
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates untrusted data from the local repository's git history and staged changes into the agent's context, which could be exploited if those sources contain malicious instructions.
- Ingestion points: The agent is instructed to read output from
git diff --cached --stat,git diff --cached, andgit logwithin SKILL.md. - Boundary markers: No specific delimiters or instructions are provided to distinguish between the repository data and the agent's core instructions.
- Capability inventory: The skill uses local git tools to inspect the project state and history.
- Sanitization: The repository output is not sanitized or filtered before being processed by the agent.
Audit Metadata