NYC

firebase-ai-logic

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS] (SAFE): The skill recommends installing firebase-tools, firebase, and @anthropic-ai/sdk. These are official packages from Google and Anthropic, which are recognized as trusted organizations. Per [TRUST-SCOPE-RULE], these downloads are considered safe.
  • [CREDENTIALS_UNSAFE] (SAFE): The code snippets use descriptive placeholders like YOUR_API_KEY rather than real secrets. The documentation explicitly warns against hardcoding keys and recommends using environment variables.
  • [PROMPT_INJECTION] (LOW): The skill implements functions that process user-provided strings and images. While this introduces a surface for indirect prompt injection (Category 8), the skill includes explicit constraints requiring input validation and prohibiting the transmission of sensitive data to the AI model, which mitigates the risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:18 PM