implement-sso
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill includes code snippets in SKILL.md (Steps 3, 4, and 5) that ingest query parameters from req.query. This constitutes a standard ingestion point for untrusted data. Boundary markers are absent in the code snippets. Capabilities involve using the SDK for authorization URL generation and token exchange. Sanitization is not explicitly implemented in the provided routes, relying on the library logic.
- [EXTERNAL_DOWNLOADS]: The skill references the @scalekit-sdk/node package. It contains explicit guardrails in SKILL.md that forbid the agent from running npm install or starting the server.
- [CREDENTIALS_UNSAFE]: The skill correctly instructs the user to store API credentials in a .env file, avoiding the risk of hardcoded secrets.
Audit Metadata