agent-md-refactor
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a logical workflow for refactoring documentation and lacks network access, code execution, or other high-risk capabilities.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process instructions from external files (e.g., AGENTS.md, CLAUDE.md), which serves as an ingestion point for untrusted data. While there are no explicit boundary markers or sanitization steps mentioned, the skill's capabilities are restricted to reading and writing markdown files within the local project directory, and it lacks high-risk tools for command execution or remote communication. 1. Ingestion points: Agent instruction files such as AGENTS.md, CLAUDE.md, and COPILOT.md (referenced in SKILL.md). 2. Boundary markers: None explicitly defined for the ingested file content. 3. Capability inventory: Performs file system read and write operations for markdown documentation files. 4. Sanitization: No specific sanitization of ingested content is performed before interpolation.
Audit Metadata