api-fuzzing
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill outlines a methodology for gathering information from external, untrusted sources such as API specifications (Swagger/OpenAPI), JavaScript bundles, and mobile application backends. If an agent follows these instructions to process data from a malicious or compromised API endpoint, that data could contain hidden instructions designed to influence the agent's subsequent actions.
- Ingestion points: Data is ingested from external API endpoints,
/swagger.json,/api-docs, and introspection results as described inSKILL.mdunder the 'Method' section. - Capability inventory: The skill suggests using tools like
ffuf,kiterunner, andarjunto probe and map external systems. - Boundary markers: The instructions do not define clear boundaries or provide guidance on how to differentiate between processed data and instructions.
- Sanitization: No sanitization or validation logic is proposed for the data retrieved from external sources before the agent processes it.
Audit Metadata