firebase-ai-logic-basics
Pass
Audited by Gen Agent Trust Hub on Apr 27, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses standard
npx -y firebase-tools@latestcommands for project initialization and listing. These are standard developer tools for the Firebase platform and are used appropriately for setup. - [EXTERNAL_DOWNLOADS]: The skill installs the official
firebaseandfirebase-toolspackages from the standard NPM registry. These are well-known services and the packages are official tools from the vendor (Firebase/Google). - [CREDENTIALS_UNSAFE]: No hardcoded credentials were found. The skill correctly advises the use of Firebase Remote Config for dynamic configuration and includes placeholders for Firebase configuration objects (
firebaseConfig), which is a safe practice. - [PROMPT_INJECTION]: No prompt injection patterns, safety bypasses, or system prompt extraction attempts were detected. The instructions are focused on guiding the developer through legitimate SDK implementation.
- [DATA_EXFILTRATION]: No evidence of data exfiltration or unauthorized network operations was found. The network calls are directed towards official Firebase/Google API endpoints for Gemini model inference.
Audit Metadata