repo-summary
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Executes repo-activity, git log, and git statistics to retrieve and analyze repository history and metrics.
- [PROMPT_INJECTION]: Employs the 'IMPORTANT' keyword to specify agent termination logic; while this is a known override marker, in this context it is used for workflow control rather than a safety bypass.
- [PROMPT_INJECTION]: Indirect prompt injection surface identified due to processing untrusted data from git logs. 1. Ingestion points: Repository commit messages and log entries (file: SKILL.md). 2. Boundary markers: None present to delimit untrusted commit data. 3. Capability inventory: Shell command execution (git) and local filesystem write operations (markdown report generation). 4. Sanitization: No specific filtering or validation of commit message content is documented.
Audit Metadata