scan
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a documentation generator that maps project structure, patterns, and conventions by scanning source code and configuration files. It does not perform any network operations or access sensitive external resources.
- [COMMAND_EXECUTION]: The skill utilizes local Node.js utility scripts located within the
.claude/hooks/lib/directory for configuration resolution and file state management. These includeproject-config-loader.cjsanddoc-stamp-guard.cjs. These are used to ensure the skill operates within the project's defined boundaries by verifying document states and loading project-specific configuration. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and summarize project source code, which constitutes an indirect prompt injection surface. This is addressed as follows:
- Ingestion points: The skill reads arbitrary project source files, build manifests, and configuration files during its scanning phases.
- Boundary markers: The skill uses structured sub-agent 'Think scopes' and mandatory section headers to separate instructions from analyzed data.
- Capability inventory: The skill has the ability to write documentation files and execute local Node.js hooks via sub-processes.
- Sanitization: The skill implements a 'Fresh-eyes' verification protocol where an independent sub-agent with zero memory of the initial scan validates all code examples and claims against the repository, ensuring the output is grounded in evidence rather than instructions found in the data.
Audit Metadata