review-spec
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill operates within its defined scope of reviewing documentation and verifying code consistency using local repository tools.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content by reading technical specification files and searching repository code via file-read and search tools. This surface is essential for its primary function.
- Ingestion points: Technical specification files specified via arguments and code files discovered via repository traversal (SKILL.md).
- Boundary markers: The prompt uses structured markdown headers to separate instructions from the document info, but does not implement strict escaping or delimiters for the raw file content being reviewed.
- Capability inventory: The skill has access to LLM invocation (mcp__codex__codex), file reading, and shell search tools (SKILL.md).
- Sanitization: No explicit sanitization or filtering of external document content is performed prior to processing.
Audit Metadata