clean-arch
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from external repositories using the
Exploresubagent. - Ingestion points: The skill performs a "Deep Architecture Scan" (Phase 2) of the repository's directory structure, server/client layers, and source files to map dependencies and logic.
- Boundary markers: Absent. There are no instructions to use delimiters or specialized prompts to prevent the agent from following instructions embedded within the scanned code (e.g., in markdown documentation or code comments).
- Capability inventory: The skill provides detailed refactoring suggestions and "Fix" examples. The "Cleanup" scope (Phase 3) implies the ability to modify the filesystem by removing dead code or merging implementations.
- Sanitization: Absent. The skill does not describe any validation or filtering of the repository content before it is processed for architectural analysis.
Audit Metadata