hive-prediction-markets

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The required workflow calls search_tools / invoke_api_endpoint to fetch prediction-market search/events/market stats from external venues at runtime, ingesting outsiders’ free-form content (e.g., market titles/descriptions or other scraped text from third-party sites/APIs) into the agent’s LLM context—an indirect prompt-injection risk.