mobile-ai-features

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFE
Full Analysis
  • [CREDENTIALS_SAFE]: The skill provides explicit warnings against bundling API keys within mobile app binaries, noting that they can be easily extracted via decompilation. It demonstrates the correct pattern of using a backend proxy (e.g., Cloudflare Workers) where secrets are stored in environment variables (env.OPENAI_API_KEY).
  • [EXTERNAL_DOWNLOADS]: All external references target official documentation and well-known services, including OpenAI, Anthropic, Google AI, and Expo. These are trusted industry-standard sources for the functionality described.
  • [COMMAND_EXECUTION]: Instructions use standard development commands (npx expo install) to add verified libraries (expo-av, expo-camera) to the project environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill documents the ingestion of user-supplied prompts and image data for AI processing. This is a standard functional surface for AI integration skills and is implemented here with appropriate authentication patterns (Bearer tokens) and architectural separation.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 07:49 AM
Security Audit — agent-trust-hub — mobile-ai-features