leadbook

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

• Third-party content exposure detected (high risk: 0.75). High likelihood: Phase 2/6 explicitly supports optional runtime providers that fetch and ingest public Web/Xiaohongshu/WeChat article/comment text into local Markdown/JSON (e.g., research/xhs/*.md / research/wxmp/...), which later gets merged into dist/book.md and thus into the agent’s LLM context for writing/quality steps.