ito-market-intelligence
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill gathers and processes information from external, untrusted sources including social media platforms, prediction market venues, and general web content. This ingestion surface is susceptible to indirect prompt injection where malicious instructions embedded in the data could influence the agent's behavior.
- Ingestion points:
SKILL.mdidentifies data gathering from external APIs (Polymarket, Kalshi), social signals (X), and general search results (Exa, GitHub, web data). - Boundary markers: The instructions do not define technical boundary markers or specific delimiters for separating untrusted data from the agent's instructions, though they do advise a conceptual separation of facts and interpretation.
- Capability inventory: The skill uses tools for web research (
deep-research,exa-search,x-api) and accesses market data via read-only API calls. - Sanitization: There are no descriptions of sanitization or escaping protocols for the external content before it is processed by the agent.
Audit Metadata