What it is

PandaWiki is an open-source knowledge base system driven by large language models, written in TypeScript and released under the AGPL-3.0 license. It lives in the self-hosted documentation and knowledge-base ecosystem, alongside the AI and LLM tooling that powers its features. The project is designed to build intelligent product documentation, technical documentation, FAQ pages, and blog systems, and it ships as a self-hosted application that runs on a Linux server through Docker. Its topic list places it squarely in the ai, docs, documentation, kb, llm, self-hosted, and wiki space, and the repository has accumulated more than ten thousand stars and over a thousand forks since it was created roughly a year ago.

The concrete problem it solves is that a traditional wiki stores documents but does not understand them. PandaWiki turns a collection of documents into a knowledge base that a language model can read, search, and answer questions from. Once documents are uploaded and learned, the system provides AI-assisted writing, AI-assisted question answering, and AI-assisted search over that content. It also removes the manual work of assembling a knowledge base by importing content from web URLs, website sitemaps, RSS feeds, and offline files, and it lets the resulting knowledge base be delivered through a hosted Wiki site, an embedded web widget, or chat bots on DingTalk, Feishu, and WeCom.

Key capabilities

• AI-assisted creation, question answering, and search, all driven by a configured large language model.
• Rich text editing that is compatible with both Markdown and HTML.
• Export of documents to Word, PDF, and Markdown formats.
• Embedding the knowledge base as a web widget on other websites.
• Delivery as chat bots for DingTalk, Feishu, and WeCom.
• Content import from a web URL, a website Sitemap, an RSS feed, or offline files.
• Separation of knowledge bases, where each knowledge base is a set of documents that generates its own Wiki site.

Who uses it and how

• Teams that need internal or public product documentation and want AI question answering layered on top of it.
• Support and community groups that publish an FAQ and let users query it through a chat bot on DingTalk, Feishu, or WeCom.
• Publishers who maintain a blog or documentation site and embed the knowledge base as a widget on an existing website.
• Operators migrating existing content by pointing the importer at a URL, a Sitemap, an RSS feed, or a set of offline files instead of retyping it.
• Self-hosters who want the whole stack on their own Linux server rather than on a vendor platform.

Getting started

Installation requires a Linux system with Docker 20.x or newer, and the README instructs running `bash -c "$(curl

bash -c "$(curl -fsSLk https://release.baizhi.cloud/panda-wiki/manager.sh)"

SUCCESS  控制台信息:
SUCCESS    访问地址(内网): http://*.*.*.*:2443
SUCCESS    访问地址(外网): http://*.*.*.*:2443
SUCCESS    用户名: admin
SUCCESS    密码: **********************