live-dev-storage-init
Pass
Audited by Gen Agent Trust Hub on Aug 8, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows security best practices by recommending that sensitive service account credentials be stored in environment variables (
FIREBASE_SERVICE_ACCOUNT) rather than being hardcoded, preventing accidental credential exposure.\n- [SAFE]: The use of the official Firebase CLI and Playwright tools for provisioning and testing is consistent with standard development workflows and represents safe tool usage.\n- [SAFE]: Project ID generation is handled using a predictable pattern based on the repository name, ensuring resource isolation in a secure manner.\n- [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes project-specific metadata (repository name) and internal reference files to generate configuration and execute CLI commands. This is standard behavior for automation scripts.\n - Ingestion points: The repository name variable and the content of
references/react-firestore-rules.md(as specified inSKILL.md).\n - Boundary markers: No explicit delimiters are used to wrap the interpolated repository name.\n
- Capability inventory: The skill performs file system writes (
firebase.json,firestore.rules) and executes shell commands through thefirebaseCLI.\n - Sanitization: There is no evidence of repository name validation or sanitization prior to its use in generated commands.
Audit Metadata