vcs-identity-check
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute standard version control CLI tools including git, hg (Mercurial), and jj (Jujutsu) to retrieve and update local user configuration for names, emails, and commit signing preferences.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from local version control configuration and commit logs which could potentially contain malicious instructions.
- Ingestion points: Identity values (name, email) and commit metadata retrieved via commands like
git configandgit login SKILL.md. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are used when the agent displays or processes these values.
- Capability inventory: The skill utilizes shell commands for reading and writing VCS configuration files (SKILL.md).
- Sanitization: No sanitization, filtering, or validation is performed on identity strings before they are incorporated into the agent's reasoning or confirmation prompts.
Audit Metadata