firebase-ai-logic
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The instructions follow established patterns for cloud service integration and AI SDK usage.\n- [EXTERNAL_DOWNLOADS]: Recommends installation of standard packages (firebase-tools, firebase, @anthropic-ai/sdk) from the official npm registry. These are well-known and trusted dependencies.\n- [PROMPT_INJECTION]: The skill defines an interface for processing user-provided prompts and images. While this creates an ingestion point for untrusted data, the skill includes explicit constraints for user input validation and the protection of sensitive data. 1. Ingestion points: prompt and imageUrl variables in SKILL.md. 2. Boundary markers: Not demonstrated in code. 3. Capability inventory: Content generation via firebase/ai. 4. Sanitization: Mandated by constraints.
Audit Metadata