crypto-risk
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process large amounts of untrusted data from multiple external web sources, creating a significant surface for indirect prompt injection attacks.
- Ingestion points: Data is collected via
WebSearchandWebFetchfrom over 10 priority sources, including CoinGecko, DeFiLlama, and various crypto news/audit sites listed in the 'Data Collection' phase. - Boundary markers: The instructions lack explicit boundary markers or directives for the agent to treat fetched content as untrusted data rather than instructions.
- Capability inventory: The skill has the capability to write the processed (and potentially poisoned) data to a local markdown file (
CRYPTO-RISK-[TOKEN].md). - Sanitization: There is no mention of sanitization, filtering, or validation of the content retrieved from external URLs before it is processed by the model.
Audit Metadata