geti-using-the-pipeline
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to fetch and interpret an OpenAPI specification from a user-provided URL (
/api/openapi.json) to determine valid endpoints and payloads. This creates a surface where a maliciously crafted schema could attempt to influence the agent's behavior. - Ingestion points: Retrieval of the API specification from
/api/openapi.jsonbased on a user-supplied base URL. - Boundary markers: The instructions do not specify any boundary markers or instructions to ignore embedded prompts within the fetched JSON schema.
- Capability inventory: The skill facilitates a wide range of actions including media uploads, project configuration, and asynchronous job management through the
GET,POST, andPATCHmethods. - Sanitization: There is no mention of validating or sanitizing the OpenAPI JSON content before the agent uses it to construct API calls.
Audit Metadata