grade-tests
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes source code from test methods and production files provided by the user. This data ingestion creates a surface where malicious instructions could be embedded in the code files to influence the agent's output. However, the skill incorporates a mandatory validation step (Step 0) that restricts processing to explicit, user-defined scopes, which serves as a significant boundary marker.
- Ingestion points: Test method bodies, production code, and diff hunks provided as input for grading (defined in SKILL.md).
- Boundary markers: Includes a mandatory validation step (Step 0) that rejects ambiguous requests and requires an explicit list of tests or files before any file reading occurs.
- Capability inventory: The skill performs file read operations to capture source code for static analysis.
- Sanitization: None mentioned for the interpolation of code content into the analysis prompt.
Audit Metadata