skills/hookdeck/agent-skills/outpost/Gen Agent Trust Hub

outpost

Pass

Audited by Gen Agent Trust Hub on Oct 1, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill's reference templates download and execute setup scripts from well-known services.
  • The Playwright Dockerfile (examples/fastapi-saas/frontend/Dockerfile.playwright) downloads the Bun installation script from bun.sh.
  • The frontend template fetches the OpenAPI schema and generates client code using standard Node.js tooling.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides a Full Stack application template that processes data from the Hookdeck Outpost API, creating an ingestion surface for potentially untrusted content.
  • The React frontend in examples/fastapi-saas/frontend/src/components/Destinations/CreateDestination.tsx and index.tsx uses dangerouslySetInnerHTML to render icon SVGs returned by the Outpost API. While restricted to the vendor's API response, this represents a point of entry for external data into the browser's DOM context.
  • The backend BFF (Backend For Frontend) routes in backend/app/api/routes/outpost.py proxy requests to the Outpost API without extensive sanitization of the forwarded response bodies, relying on the client-side framework (React/TypeScript) for presentation-layer safety.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 1, 2026, 03:49 PM