What problem does it solve? Academic writing in Chinese requires authentic, traceable references from CNKI (知网), but AI models often fabricate citations from memory. This Skill enforces a strict verification workflow so every Chinese reference comes from real CNKI page data or exported citations, never from hallucinated metadata. ## Core Features & Use Cases - Browser-task driven CNKI retrieval: Delegates searching, login checks, filtering, and batch GB/T 7714 citation export to a browser task, with clear rules for login, CAPTCHA, and institutional authentication handling. - Citation pool and evidence map: Organizes verified references into a CNKI citation pool and an evidence map linking each source to supported claims before any writing begins. - Failure transparency: Distinguishes verified from unverified references, forbids substituting scholar_search or API metadata for page-visible results, and provides manual fallback templates when browser tasks are unavailable. - Use Case: When writing a thesis introduction or literature review, the Skill first searches CNKI for 10-20 candidate papers, exports their GB/T 7714 citations, builds an evidence map, and only then drafts the cited text. ## Quick Start Ask the assistant to search CNKI for recent core-journal papers on your thesis topic and export verified GB/T 7714 references before writing the introduction.
