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

outpost

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes a secure BFF (Backend-for-Frontend) pattern in backend/app/api/routes/outpost.py. This architecture ensures that sensitive administrative API keys (OUTPOST_API_KEY) are managed entirely on the server and are never exposed to the client-side browser environment.
  • [SAFE]: Tenant isolation is programmatically enforced in the API routes. The application maps authenticated users to isolated tenants using their unique user IDs as the tenant_id, preventing cross-tenant data access or unauthorized event delivery.
  • [SAFE]: The provided FastAPI reference template includes proactive security measures, such as constant-time comparison logic in backend/app/crud.py to prevent timing attacks during authentication, and automated checks in backend/app/core/config.py that block deployment if default placeholder credentials are used.
  • [SAFE]: All external downloads and remote scripts, such as the Bun installation in examples/fastapi-saas/frontend/Dockerfile.playwright, originate from well-known technology services and official domains.
  • [SAFE]: Standard security practices for web applications are followed throughout the examples, including JWT-based session management, secure CORS configuration, and appropriate input validation using Pydantic and Zod.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 12:01 AM
Security Audit — agent-trust-hub — outpost