firebase
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes patterns for ingesting data from external sources like Firestore and Firebase Authentication, which constitutes an injection surface.
- Ingestion points:
docSnap.data()andsnapshot.data()inreferences/detailed-guide.md, and HTTP request headers in the Cloud Functions examples. - Boundary markers: The skill explicitly instructs on using Firestore Security Rules and server-side token verification as security boundaries.
- Capability inventory: The skill proviｄes code for database interactions and server-side logic execution via Cloud Functions.
- Sanitization: The guiｄe emphasizes verifying ID tokens and using access control rules to valiｄate user identity.
- [EXTERNAL_DOWNLOADS]: The skill references standard, well-known libraries for Firebase integration.
- Evidence: References to Node.js packages such as
firebase,firebase-admin,firebase-functions,reactfire, andangularfire, which are official or wiｄely-useｄ tools for the Firebase platform.
Audit Metadata