claude-cockpit
MCP server for a kanban task board that lets Claude Code manage tasks with priority, categories, and multi-agent safety, plus a web UI for visual drag-and-drop.
README
Claude Cockpit
A kanban task board with an MCP server and web UI, designed for managing tasks with Claude Code.
Claude agents can pick tasks, claim them, move them across columns, and work through a prioritized backlog — all through MCP tools. A web dashboard lets you drag and drop tasks, view details, and manage the board visually.
Features
- Kanban board with columns: bank, todo, in_progress, testing, done
- MCP server — Claude Code can list, create, claim, move, and update tasks
- Web UI — drag-and-drop board at
http://localhost:3333with live WebSocket updates - Multi-agent safe — file-based locking prevents race conditions when multiple agents work concurrently
- Claude Code plugin — slash commands (
/board,/pick,/done,/create) and hooks for automated workflow - Snooze app integration — optional integration with the Snooze task app (calendar import, task sync)
- Task priorities & categories — prioritize work and tag tasks by type
- Image attachments — attach screenshots and images to tasks
- AI auto-categorization — tasks are auto-tagged using Claude when created via MCP
Prerequisites
- Node.js 20+
- Claude Code CLI (install guide)
- jq (for shell hooks) —
brew install jqon macOS
Installation
# Clone the repo
git clone https://github.com/shmuelix/claude-cockpit.git
cd claude-cockpit
# Install dependencies
npm install
# Build
npm run build
Setup
1. Environment variables
Copy the example env file and fill in your keys:
cp .env.example .env
Edit .env:
OPENAI_API_KEY— required for AI auto-categorization of tasksCALENDAR_FEED_URL— optional, for Snooze calendar import
2. Start the web server
npm run web
This starts the web dashboard at http://localhost:3333.
3. Configure Claude Code MCP server
Add this to your Claude Code MCP settings (~/.claude/plugins/claude-cockpit.json or via claude mcp add):
claude mcp add claude-cockpit -- node /absolute/path/to/claude-cockpit/build/mcp-server.js
Or manually add to your Claude Code settings:
{
"mcpServers": {
"claude-cockpit": {
"command": "node",
"args": ["/absolute/path/to/claude-cockpit/build/mcp-server.js"]
}
}
}
4. Install the Claude Code plugin (optional but recommended)
This gives you slash commands and workflow hooks:
claude plugin add /absolute/path/to/claude-cockpit/plugin
This enables:
/board— show board status/pick— pick the next task/done— finish current task and pick next/create— create a new task- Auto-continue workflow after completing tasks
- Terminal title updates showing current task
5. Snooze app integration (optional)
If you use the Snooze app, copy the example config and fill in your credentials:
cp snooze-config.example.json snooze-config.json
Edit snooze-config.json with your Snooze userId and API key.
Usage
Web UI
Open http://localhost:3333 after starting the web server. You can:
- Drag and drop tasks between columns
- Click tasks to view/edit details
- Create new tasks with the + button or press
n - Add notes and image attachments
- Filter by category and priority
With Claude Code
Once the MCP server is configured, Claude Code can manage your board:
> /pick # Pick the next task and start working
> /board # Show board status
> /done # Finish current task
> /create Fix the login bug on mobile
Claude will automatically:
- Claim tasks before working on them
- Add notes as it works
- Move tasks to testing/done when finished
- Pick the next task automatically
MCP Tools
Available tools when connected via MCP:
list_tasks— list all tasks grouped by columnget_task— get full task detailscreate_task— create a new taskupdate_task— update task fieldsmove_task— move task between columnsclaim_task/unclaim_task— claim/release task ownershipadd_note— add a note to a taskdelete_task— delete a tasksnooze_*— Snooze app integration tools (if configured)
Project Structure
claude-cockpit/
├── src/
│ ├── mcp-server.ts # MCP server (Claude Code interface)
│ ├── web-server.ts # Express + WebSocket web server
│ ├── store.ts # Task storage with file-based locking
│ ├── types.ts # TypeScript types
│ ├── snooze-api.ts # Snooze app API client
│ └── import-todo.ts # Import tasks from external sources
├── public/
│ └── index.html # Web UI (single-page app)
├── plugin/
│ ├── commands/ # Claude Code slash commands
│ ├── hooks/ # Workflow automation hooks
│ └── skills/ # Claude Code skills
├── tasks.json # Task data (auto-created, gitignored)
├── column-orders.json # Column sort order (auto-created, gitignored)
└── uploads/ # Image attachments (gitignored)
Data
All task data is stored locally in tasks.json. No external database required. The file is created automatically when you first create a task.
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.