firebase-firestore

Pass

Audited by Gen Agent Trust Hub on Aug 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses npx -y firebase-tools@latest to list, create, and manage Firestore databases. It also employs curl to fetch Maven metadata and executes local gradlew commands to inspect project dependencies. These operations use standard, well-known development tools.
  • [EXTERNAL_DOWNLOADS]: The skill installs the firebase-tools CLI from the official NPM registry and downloads SDK version metadata from Google's Maven repository (dl.google.com). These sources are trusted technology providers.
  • [PROMPT_INJECTION]: The skill contains strict instructional overrides (using terms like 'STRICTLY FORBIDDEN', 'MANDATORY', and 'CRITICAL') to enforce specific architectural patterns and security protocols. It also directs the agent to adopt an 'Expert Firebase Security Rules engineer' persona and follow a rigid three-phase workflow for security analysis, which includes scanning the local codebase for data models and access patterns.
  • [DATA_EXPOSURE]: To facilitate security analysis, the instructions guide the agent to scan the entire local codebase for Firestore-related queries, schemas, and authentication patterns. This is part of the intended functionality for generating project-specific security rules.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 14, 2026, 01:44 PM
Security Audit — agent-trust-hub — firebase-firestore