document-summary
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE] (SAFE): No malicious patterns were identified in the skill instructions or metadata.
- [NO_CODE] (SAFE): The skill is comprised solely of natural language documentation and lacks any executable scripts or binary files.
- [PROMPT_INJECTION] (LOW): The skill has a surface for indirect prompt injection. 1. Ingestion points: External document content. 2. Boundary markers: Not defined. 3. Capability inventory: No executable tools or network access. 4. Sanitization: None. Because the primary purpose is summarization and there are no high-risk capabilities, the risk is negligible.
Audit Metadata