1. Project Overview

ChatWiki is a workflow automation platform focused on WeChat ecosystem, committed to making every public account a super AI intelligent body. It integrates the open capabilities of the public account platform, and can be built into a WeChat ecosystem application by dragging and dropping. It can implement public account tweet one-click editing, message AI selected reply, etc.

2. Open Source License

Licensed under Apache-2.0

3. Screenshots

System Screenshot：

4. Feature Overview

ChatWiki is a workflow automation platform focused on WeChat ecosystem. It integrates the open capabilities of the public account platform, and can be built into a WeChat ecosystem AI intelligent body by dragging and dropping. It can implement public account tweet one-click editing, message AI selected reply, etc. It can build a private AI knowledge base in 5 minutes to quickly construct a customer service robot. It is committed to becoming a vertical field coze, n8n.

Core Features

WeChat ecosystem deep integration

• Industry first
  Unverified public account private message automatic reply, supports text, voice, image, small program card, video message, etc.
• WeChat workflow
  Integrate user private message, message, follow, unfollow, click menu, etc. trigger scenarios, support reply private message, fan label, generate draft article, publish article, etc. multiple processing flows
• Knowledge base synchronization
  Supports grabbing public account article materials, one-click establishing knowledge base

Basic capabilities

• Workflow orchestration:
  Dialogue workflow, plugin workflow, including basic workflow nodes, double-sided MCP, Agent mode, user interaction
• Document knowledge base:
  Supports URL reading, document batch import, API interface, supports AI segmentation, QA segmentation, parent-child segmentation. Supports knowledge graph, vector mixed search, can visualize knowledge graph
• Question and answer knowledge base:
  Upload documents to automatically extract question and answer knowledge, supports unknown questions automatic clustering, supports summarizing common FAQs from human dialogue
• Transfer human customer service:
  Handle general user inquiries through robots, while supporting human customer service reception. Robot handling of bad questions can be intervened by human customer service, supports multiple customer service collaborative distribution
• Model support:
  Supports DeepSeek R1, doubao pro, qwen max, Openai, Claude, etc. more than 20 mainstream models worldwide

More capabilities

• Multiple deployment methods
  Provide desktop client, support publishing as WebApp, support embedding website, public account service number, WeChat customer service, WeChat small shop customer service, etc.
• MCP & API integration
  Can introduce external MCP services, or publish workflow as MCP service. Complete OpenAPI interface, easy to integrate existing business systems
• Multi-account permission management
  Manage, edit, view three-level permission system, implement data permission isolation. IP whitelist, login log permanent retention

5. Tech Stack

One-click deployment

ChatWiki community edition is based on Docker deployment, only a few steps are needed to complete the installation:

# Install Docker
sudo curl -sSL https://get.docker.com/ | CHANNEL=stable sh
# Clone the project
git clone chatwiki.gitcd chatwiki/docker
# Start Service
docker compose up -d
# Start using, access by IP + port (need to open specified port ${CHAT_SERVICE_PORT}, default 18080)
# Default account: admin
# Default password: chatwiki.com@123

Tech Stack

• Front-end: vue.js
• Back-end: golang + python
• Database: PostgreSQL16 + pgvector + zhparser