Self-hosted enterprise knowledge hub that compiles company docs into a wiki and serves it to Claude and other LLMs via an MCP server with department-scoped access.
1.5K297PythonOtherUpdated 5 months ago
Tencent/WeKnora
Open-source LLM knowledge platform from Tencent that turns documents into a RAG Q&A system, a reasoning agent, and an auto-maintained wiki.
Open-source alternative to:Notion AI
- Stars
- 35,086
- Forks
- 4,704
- Open issues
- 773
- Last commit
- 22 hours ago
What is it?
WeKnora is an LLM-powered knowledge framework for enterprise document understanding, semantic retrieval and reasoning. It gathers a team's documents into knowledge bases that can be searched with RAG, worked on by a multi-step agent, or organized into an interlinked wiki with a knowledge graph.
The agent can call knowledge base search, web search, MCP tools and skills, and it can run skills in session-persistent Docker, E2B or Cube sandboxes. Documents can be synced from Feishu, Confluence, GitLab, Notion, Yuque and other sources, and parsed from formats including PDF, Word, Excel and XMind. Retrieval combines keyword and vector search with reranking and parent-child chunking.
Models, vector databases and object storage backends are swappable, with 29 built-in LLM vendors and Ollama support. It deploys via Docker Compose, Helm on Kubernetes, or as a Lite single binary using SQLite, and includes workspace RBAC, API keys, and Langfuse tracing.
Key features
- 01RAG Q&A with hybrid keyword and vector search, reranking and source citations
- 02Multi-step ReAct agent that uses knowledge bases, web search, MCP tools and skills
- 03Auto-generated wiki pages with a knowledge graph, editing and rollback
- 04Data source sync from Feishu, Confluence, GitLab, Notion and other platforms
- 05Swappable LLMs, vector databases and object storage backends
- 06Workspace RBAC, scoped API keys and Langfuse tracing
Use cases
- Building an internal Q&A assistant over company manuals, policies and engineering docs
- Letting staff query scattered knowledge from Feishu, Confluence or Notion in one place
- Publishing a knowledge base to Cursor or Claude through the built-in MCP server
- Running a private, self-hosted knowledge system on an internal network
Quick start
- 1
Clone the repository
git clone https://github.com/Tencent/WeKnora.git - 2
Enter the project folder and create the environment file
cd WeKnora && cp .env.example .env - 3
Pull the latest images
docker compose pull - 4
Start the core services, then open http://localhost
docker compose up -d
Things to know
- Full deployment needs Docker, Docker Compose and several services, so self-hosting takes setup effort.
- The README warns that production deployments should stay on a private network, not the public internet.
- Some features need external model API keys or optional services such as Neo4j, MinIO or Langfuse.
In collections
Topics
FAQ
- What is WeKnora?
- Open-source LLM knowledge platform from Tencent that turns documents into a RAG Q&A system, a reasoning agent, and an auto-maintained wiki. WeKnora is an LLM-powered knowledge framework for enterprise document understanding, semantic retrieval and reasoning. It gathers a team's documents into knowledge bases that can be searched with RAG, worked on by a multi-step agent, or organized into an interlinked wiki with a knowledge graph.
- Is WeKnora free to use? What license does it have?
- WeKnora is open source under the Other license. Check the license terms before using it in commercial products.
- Is WeKnora still maintained?
- Actively maintained. The last commit to WeKnora was on Oct 22, 2026.
- What can I use WeKnora for?
- Building an internal Q&A assistant over company manuals, policies and engineering docs Letting staff query scattered knowledge from Feishu, Confluence or Notion in one place Publishing a knowledge base to Cursor or Claude through the built-in MCP server Running a private, self-hosted knowledge system on an internal network
- How do I install WeKnora?
- Clone the repository: git clone https://github.com/Tencent/WeKnora.git Enter the project folder and create the environment file: cd WeKnora && cp .env.example .env Pull the latest images: docker compose pull Start the core services, then open http://localhost: docker compose up -d
- Where can I get WeKnora?
- The source code and installation instructions are on GitHub: https://github.com/Tencent/WeKnora
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Self-hostable AI second brain that answers questions from the web and your own documents, with custom agents and scheduled automations.
37.8K2.5KPythonAGPL-3.0Updated 3 months ago
A local-first Markdown memory workspace that lets AI agents store, search, and consolidate long-term knowledge as editable files.
3.7K320PythonApache-2.0Updated yesterday
Self-hostable personal AI assistant as a single Go binary, combining LLM chat, RAG knowledge bases, and agent loops for browsing, shell, and MCP tools.
5.7K667GoApache-2.0Updated yesterday
An open-source LLM application devops platform for enterprises, covering workflows, RAG, agents, model management, evaluation and fine-tuning.
12.1K2KPythonApache-2.0Updated 21 hours ago
Memory operating system for LLMs and AI agents that stores, retrieves, and manages long-term memory as an editable graph.
12K1.1KTypeScriptApache-2.0Updated 2 days ago