cleanup
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted repository content (source, documentation, and configuration) which constitutes an indirect prompt injection surface.
- Ingestion points: The agent reads target repository files and project instructions during the scope establishment phase defined in "SKILL.md".
- Boundary markers: The instructions explicitly mandate excluding sensitive or ignored material and require an explicit user request before execution.
- Capability inventory: The skill is capable of performing file writes to apply authorized changes and executing subprocess workers using existing system tooling.
- Sanitization: No specific data sanitization or escaping of the processed repository content is implemented.
- [COMMAND_EXECUTION]: The skill facilitates the use of subprocesses to delegate tasks to existing system tools.
- Evidence: "SKILL.md" provides guidance on using "subprocess workers" when suitable tooling is already available and permitted, while explicitly forbidding the installation of new tools just for this purpose.
Audit Metadata