coingecko
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from the external CoinGecko API, which serves as a vector for indirect prompt injection through unverified metadata such as token descriptions.
- Ingestion points: Data is ingested from the CoinGecko on-chain API endpoints as demonstrated in
templates/coingecko-client.tsandexamples/integration/full-client.ts. - Boundary markers: There are no explicit boundary markers or instructions provided in the templates to differentiate between trusted and untrusted content within API responses.
- Capability inventory: The integration is limited to network requests (
fetch) and JSON parsing. It does not exhibit capabilities for file writing, subprocess execution, or dynamic code evaluation based on the ingested data. - Sanitization: The implementation templates lack sanitization for string-based fields returned by the API, which could potentially contain malicious instructions.
Audit Metadata