[KongchangAI]
Product微信知识库

WeKnora

A local knowledge base tool open-sourced by Tencent's WeChat team. It supports RAG question answering, task execution, MCP, skills, and client features, can generate Wiki pages with knowledge links, and can be invoked as a component by AI tools such as Claude Code.

Timeline (last 90 days)

Oct 28

WeKnora supports knowledge Q&A, task execution, MCP, skills, clients and other features, and can generate Wiki pages with knowledge links, and can be called as a component by AI tools such as Claude Code.

Unverified50%
Oct 28

WeKnora officially recommends containerized deployment, but the standard method has relatively high memory requirements (about 4-8GB).

Unverified50%
Oct 28

The official WeKnora documentation lists three deployment forms: production environment (containerized/K8s), learning environment (binary), and desktop application (not yet released).

Unverified50%
Oct 28

The WeKnora learning environment uses a file-based database and in-memory storage, requires no additional dependencies, and is ready to use after unzipping.

Unverified50%
Oct 28

The default listen address configured for the WeKnora learning environment is localhost:8215.

Unverified50%
Oct 28

WeKnora's data encryption key corresponds one-to-one with the data directory, and changing the .env file or clearing the data may cause decryption to fail.

Unverified50%
Oct 28

The current binary version of WeKnora has no logout button, but you can switch accounts by clearing cookies through the browser developer tools.

Unverified50%

All Facts (8)

Source Articles