expo-api-routes
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- PROMPT_INJECTION (LOW): The skill provides examples of endpoints that ingest untrusted user data and forward it to external LLMs (OpenAI) without sanitization or boundary markers, creating a vulnerability surface for indirect prompt injection. \n
- Ingestion points: User input is ingested via
request.json()inapp/api/ai+api.tsand via query parameters inapp/api/weather+api.ts. \n - Boundary markers: The code examples do not demonstrate the use of delimiters or instructions to ignore embedded commands. \n
- Capability inventory: The skill routes use
fetchto reach external LLM and weather APIs. \n - Sanitization: No input sanitization or validation logic is included in the AI integration snippets. \n- SAFE (SAFE): The skill demonstrates best practices for managing server-side secrets using environment variables and provides legitimate instructions for using official deployment tools like
eas-cli.
Audit Metadata