NYC

firebase-ai-logic

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): Analysis of the skill instructions and code snippets confirms no malicious intent, hidden commands, or data exfiltration. The content is strictly limited to Firebase integration documentation.
  • [EXTERNAL_DOWNLOADS] (LOW): Recommends installing standard, trusted npm packages including 'firebase-tools', 'firebase', and '@anthropic-ai/sdk'. These originate from reputable organizations (Google and Anthropic).
  • [CREDENTIALS_UNSAFE] (SAFE): The skill uses clear placeholders for configuration values and includes specific 'MUST' constraints and best practices advising against hardcoding API keys.
  • [PROMPT_INJECTION] (LOW): As a guide for AI features, it describes surfaces for processing user input but explicitly includes a constraint to perform input validation to mitigate injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:08 PM