setting-up-a-custom-rest-source

Pass

Audited by Gen Agent Trust Hub on Jul 21, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill interacts with external content, such as API documentation and live JSON responses, which represents a potential indirect prompt injection surface. However, the workflow includes specific validation and preview steps that focus on structural data mapping, reducing the risk of unintended instruction execution from external data.
  • [DATA_EXFILTRATION]: The skill manages sensitive API credentials using a secure design pattern. It prohibits inlining secrets within the manifest and instead uses dedicated payload keys that interact with the platform's server-managed credential store, ensuring secrets are not exposed in the configuration or logs.
  • [COMMAND_EXECUTION]: The skill uses specific platform tools for configuration and validation. The manifest-based approach limits interaction to standard REST methods (GET/POST) and includes documented SSRF guards that prevent access to internal or private network addresses.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 21, 2026, 08:12 PM
Security Audit — agent-trust-hub — setting-up-a-custom-rest-source