sumsub-analyze-regulation
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external documents (PDF or text) to extract regulatory requirements and map them to platform capabilities, creating a data ingestion surface.
- Ingestion points: Regulation documents, legal acts, or compliance documents provided in Step 1.
- Boundary markers: The skill instructions require a mandatory validation step (Step 3: 'My Understanding') where the agent must present its findings and wait for user confirmation before generating a final configuration plan.
- Capability inventory: The skill generates text-based planning descriptions; it is explicitly prohibited from generating JSON payloads or performing direct API calls.
- Sanitization: Relies on the human-in-the-loop confirmation process to verify the extracted requirements.
- [SAFE]: The skill's behavior is consistent with its stated purpose of regulatory analysis. It references standard Sumsub ecosystem tools and utilizes the 'Read' tool as restricted in its metadata. No evidence of obfuscation, credential theft, or unauthorized persistence was detected.
Audit Metadata