implementation-contract
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The workflow involves agents reading external tasks, requirements, and evidence, which creates a potential surface for indirect prompt injection if those sources are malicious. Ingestion points: Reading tasks, requirements, and pointers to evidence (SKILL.md). Boundary markers: Instructions emphasize working within user-confirmed specifications and project norms (SKILL.md). Capability inventory: The skill provides no executable code or tools. Sanitization: No data sanitization measures are specified.
- [SAFE]: No malicious patterns were detected. The skill includes hardening measures such as disabling model invocation and preventing implicit activation.
Audit Metadata