payment-integration
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [Data Exposure] (SAFE): The skill explicitly mandates that sensitive card data never be logged and highlights the importance of using tokenization to avoid handling raw payment information.
- [Indirect Prompt Injection] (LOW): The skill processes untrusted external data via webhooks. Evidence Chain: (1) Ingestion points: Webhook request bodies and payment buttons (SKILL.md). (2) Boundary markers: None specified. (3) Capability inventory: Code generation for integration and database schemas. (4) Sanitization: Mandatory instructions for signature verification and server-side status validation.
- [Unverifiable Dependencies] (SAFE): Recommends using official, well-known SDKs for payment processing and provides checklists for secure configuration.
- [Prompt Injection] (SAFE): No malicious instructions or bypass attempts were found in the skill body or metadata.
Audit Metadata