CoCursor

Team AI Collaboration Tool for Cursor IDE

中文文档 • Releases • VS Code Marketplace

---

The efficiency gap between those who can use AI and those who can't is 100x. This is not an exaggeration.

What is CoCursor?

CoCursor is a VS Code/Cursor extension that empowers teams to collaborate with AI more effectively. It combines work analytics, semantic search of AI conversations, skill sharing marketplace, and automated reporting — all running locally with complete data privacy.

Tech Stack:

• Backend: Go 1.24 + Gin + DDD Architecture
• Frontend: VS Code Extension + React + TypeScript
• Team Collaboration: P2P Architecture + mDNS Discovery + WebSocket Real-time Sync
• RAG: Qdrant Vector Database + Embedding Models (supports local deployment)
• Workflow: OpenSpec-driven Development

Features

📊 Work Analysis Dashboard

Track every AI collaboration session automatically.

Feature	Description
Session Tracking	Monitor your work sessions in Cursor with detailed statistics
Code Analytics	Track lines added/removed, files changed, token usage trends
Time Heatmap	Visualize when you're most productive
Top Files	See which files you work on most with AI
One-Click Reports	Generate daily/weekly work reports instantly

No more spending 30 minutes writing work reports. AI helps you work and helps you report.

🔍 AI Conversation Semantic Search (RAG)

Every question, code snippet, and solution you've discussed with AI is in your Cursor chat history.

Feature	Description
Automatic Indexing	Index all your Cursor conversations locally
Semantic Search	Find conversations using natural language, not keywords
Knowledge Retrieval	"How did I solve that database issue?" → Found instantly
Project Filtering	Search within specific projects
Context Preview	See relevant context before opening full conversation

Your AI conversations are no longer one-time use — they become searchable, reusable knowledge.

🛒 Skill Marketplace

One person knowing AI isn't enough — the whole team needs to know.

Feature	Description
Browse Skills	Discover productivity-boosting AI skills
One-Click Install	Install skills directly to your Cursor configuration
Category Filters	Find skills by category (productivity, creative, tools, etc.)
Source Filters	View built-in skills or team-shared skills
Team Publishing	Share your custom skills with teammates

Let the weakest member on the team use the strongest member's AI skills.

👥 Team Collaboration

Collaborate with your team in real-time, completely within your LAN.

Feature	Description
P2P Discovery	Auto-discover team members via mDNS
Code Sharing	Right-click to share selected code with team
Daily Reports	View team members' work summaries
Weekly Calendar	See team activity at a glance
Member Stats	Track team productivity metrics

P2P LAN direct transfer — no server involved, data stays secure.

⚡ Workflow Engine

Drive AI workflows with OpenSpec specifications:

• Requirements → Design → Implementation, standardized end-to-end
• Not "what do you think we should do" but "everyone follows this process"
• AI executes according to specs, results are predictable

🔔 Daily Summary Reminder

Never forget to summarize your work.

Setting	Default	Description
Evening Reminder	17:50	Get notified before leaving work
Morning Follow-up	09:00	Reminder next morning if you missed yesterday
Enable/Disable	Off	Toggle reminders in settings

Installation

From VS Code Marketplace

Search for "CoCursor" in the VS Code/Cursor Extensions marketplace and install.

From GitHub Releases

1. Download the VSIX file for your platform from Releases:

   • cocursor-linux-x64.vsix - Linux x64
   • cocursor-win32-x64.vsix - Windows x64
   • cocursor-darwin-x64.vsix - macOS Intel
   • cocursor-darwin-arm64.vsix - macOS Apple Silicon

2. Install in VS Code/Cursor:

   code --install-extension cocursor-<platform>.vsix

Build from Source

# Clone the repository
git clone https://github.com/toheart/cocursor.git
cd cocursor

# Build backend (requires Go 1.24+)
cd backend
make build-all

# Build frontend extension (requires Node.js 18+)
cd ../co-extension
npm install
make build

# Package as VSIX
npx @vscode/vsce package

Quick Start

1. Open CoCursor Panel: Click the CoCursor icon in the VS Code/Cursor sidebar
2. Work Analysis: View your AI collaboration statistics and generate reports
3. RAG Search: Search through your historical AI conversations (requires setup)
4. Skill Marketplace: Browse and install productivity-boosting AI skills
5. Team Collaboration: Create or join a team to share skills and code

Commands

Command	Description
CoCursor: Open Dashboard	Open work analysis dashboard
CoCursor: Open Sessions	View recent AI conversation sessions
CoCursor: Open Marketplace	Browse and install AI skills
CoCursor: Share Code to Team	Share selected code with team members
CoCursor: Toggle Status Sharing	Enable/disable work status sharing
CoCursor: Refresh Webview	Refresh the CoCursor panel data

Configuration

Setting	Default	Description
cocursor.autoStartServer	true	Auto-start the backend daemon
cocursor.daemon.port	19960	Backend server port
cocursor.reminder.enabled	false	Enable daily summary reminders
cocursor.reminder.eveningTime	17:50	Evening reminder time (HH:mm)
cocursor.reminder.morningTime	09:00	Morning follow-up time (HH:mm)

RAG Setup (Optional)

To enable semantic search of your AI conversations:

1. Open CoCursor sidebar → RAG Search → Settings (gear icon)
2. Configure embedding model (supports OpenAI, local models via Ollama)
3. Set up Qdrant vector database (can run locally via Docker)
4. Click "Start Indexing" to index your conversations

Recommended Setup:

# Run Qdrant locally
docker run -p 6333:6333 qdrant/qdrant

Architecture

cocursor/
├── backend/                 # Go Backend Daemon (DDD Architecture)
│   ├── cmd/                 # Application entry points
│   ├── internal/
│   │   ├── domain/          # Domain models and business logic
│   │   ├── application/     # Application services
│   │   ├── infrastructure/  # External integrations (Qdrant, SQLite, etc.)
│   │   └── interfaces/      # HTTP handlers, MCP tools
│   └── pkg/                 # Shared packages
├── co-extension/            # VS Code Extension (React + TypeScript)
│   ├── src/
│   │   ├── extension.ts     # Extension entry point
│   │   ├── webview/         # React UI components
│   │   │   ├── components/  # WorkAnalysis, RAGSearch, Marketplace, Team...
│   │   │   ├── services/    # API service layer
│   │   │   └── hooks/       # React hooks
│   │   └── daemon/          # Daemon process manager
│   └── resources/           # Static assets
└── openspec/                # OpenSpec specifications

Privacy & Security

• 100% Local Execution: All data processing happens on your machine
• No Cloud Services: Your code and conversations never leave your computer
• P2P Team Collaboration: Direct peer-to-peer communication within your LAN
• Open Source: Fully auditable codebase
• No Telemetry: We don't collect any usage data

Roadmap

Phase	Capability	Value
Now	Personal conversation search	Personal knowledge retention
Next	MCP Integration	Connect more data sources
Future	Team Brain	Aggregate all team members' AI conversations into a team knowledge base

Imagine: A new team member doesn't need to ask veterans — just search the Team Brain: "What pitfalls did we encounter with this module?" — Everyone's experience is right there.

When every AI conversation becomes searchable knowledge, "knowledge lost when people leave" is solved forever.

Contributing

Contributions are welcome! Please read our contributing guidelines before submitting PRs.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

License

CoCursor Non-Commercial License - Free for non-commercial use only.

Links

• GitHub: https://github.com/toheart/cocursor
• VS Code Marketplace: https://marketplace.visualstudio.com/items?itemName=tanglyan-cocursor.cocursor
• Issues: https://github.com/toheart/cocursor/issues

---

If you're also leading a team and thinking about how to help your team use AI better — let's connect!