Overview
Guanlan is a Chinese-internet research CLI / MCP tool for AI agents. It helps agents with Chinese web search, source routing, page reading, hot-list watching, research evidence packs, and local knowledge-base capture. Guanlan defaults to local, read-only, low-disturbance operation—it does not actively read cookies, keychains, or login state—and fits Chinese-web retrieval, source verification, industry research, sentiment watching, and local LLM networking. Good for: Chinese-web research across policy, industry, finance, culture, tech communities, and public social pages; agent networked research that turns hits into context with sources, boundaries, and evidence roles; page reading that converts public pages to model-friendly Markdown; hot lists from Baidu, Weibo, Bilibili, V2EX, IThome, and more; local LLM networking via CLI / MCP for LM Studio, Ollama, Open WebUI, and similar. Defaults stay local, read-only, low-disturbance—no cookies/keychain/login reads, no post/like/comment/DM writes. It does not make final judgments for users; it helps agents find better Chinese sources, see source identity, and organize materials into evidence packs for further reasoning.
Connected profiles
More Intelligent search products
Sources · 0Sources & provenance
English text is maintained separately from the original record. Consult the original source before citing a fact.
No source links are available for this profile.
Original source text
Names may use phonetic renderings when no English name is supplied. The original wording is retained below.
观澜 Guanlan 是一个面向 AI Agent 的中文互联网研究 CLI / MCP 工具。它帮助 Agent 进行中文网页搜索、信源路由、网页阅读、热榜观察、研究证据包整理和本地知识库沉淀。观澜默认本地运行、只读、低扰，不主动读取 Cookie、钥匙串或登录态，适合用于中文互联网检索、资料核验、行业研究、舆情观察和本地大模型联网。它适合这些场景：- 中文互联网资料检索：政策、产业、财经、文娱、技术社区、社交公开页等- Agent 联网研究：把搜索结果整理成带来源、边界和证据角色的上下文- 网页阅读：将公开网页转成更适合模型阅读的 Markdown- 热榜观察：查看百度、微博、B站、V2EX、IT之家等公开热榜- 本地大模型联网：通过 CLI / MCP 给 LM Studio、Ollama、Open WebUI 等模型提供检索能力观澜默认本地运行、只读、低扰，不主动读取 Cookie、钥匙串或登录态，不执行发帖、点赞、评论、私信等写操作。它不是替用户做最终判断，而是帮助 Agent 找到更合适的中文信源、看清来源身份，并把材料组织成可继续推理的证据包。