rq-idea-generation

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly allows and documents using the web_search tool to collect real-time public news/policy/company items (see SKILL.md Step 4.5 and references/web_search.md) which must be written to web_search_findings.json and read by the LLM before generating summaries, so untrusted third‑party content is ingested and can materially influence report text and prioritization.