edge-hint-extractor

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can enter the LLM context via --llm-ideas-cmd: build_hints.py reads market_summary.json, anomalies.json, and especially news_reactions.csv/.json (outsider-authored public/news text) and passes their contents as JSON on stdin to an external LLM CLI, which then generates hints that are parsed back into the agent’s context.