Supercoordination

You are running a cooperative MCP server designed for cross-team collaboration within the same building or local network. It coordinates members, tasks, and team dashboards, with persistent data storage and a straightforward client workflow to keep everyone aligned.

How to use

You will connect an MCP client to the server, start the server, and then register members, create tasks, and monitor progress through the built-in dashboard. The server provides a local access point, simple configuration, and automatic data persistence to keep your team activities synchronized.

How to install

Prerequisites: Node.js (and npm) installed on your machine.

# 1. Start the server host (preferred, prevents sleep)
cd ~/ClaudeWorkspace/supercoordination-mcp
./start-server.sh

# 2. Alternative start
npm start

Configuration and usage details

Server addresses: local: http://localhost:3000 and local network: http://192.168.1.3:3000. The MCP endpoint is located at http://192.168.1.3:3000/mcp.

Team members connect by creating a working directory, configuring the MCP connection, and launching Claude Code to interact with the server. A representative setup is shown below.

# 1. Create a working directory for the MCP client
mkdir -p ~/ClaudeWorkspace/supercoordination-mcp
cd ~/ClaudeWorkspace/supercoordination-mcp

# 2. Configure MCP connection
cat > .mcp.json << 'EOF'
{
  "mcpServers": {
    "supercoordination": {
      "type": "http",
      "url": "http://192.168.1.3:3000/mcp"
    }
  }
}
EOF

# 3. Launch Claude Code
claude

# 4. Verify the connection from Claude by listing members
列出所有成员

完整闭环测试场景

Practice scenario: develop a virtual IP module for an AIGC Studio. Steps include member registration, task creation, smart matching, task allocation, progress updates, dashboard review, and balance checks.

测试命令

In Claude Code, execute the following actions to simulate a workflow.

> "Register new member: Zhang San, skills including Python, AI development, system architecture, five-element profile: Wood40%, Water30%..."

> "Register new member: Li Si, skills including content creation, brand planning, AI tools, five-element profile: Fire35%, Wood25%..."

> "Create task: Develop AIGC Studio virtual IP generation module, description: fully automate from character to content, five-element attribute: Wood, priority S, required skills Python, AI, system architecture"

> "Create task: Design virtual IP content strategy, description: define five virtual IP personas and content direction, five-element attribute: Fire, priority A, required skills content creation, brand planning"

> "Find best members for all unassigned tasks"

> "View team collaboration dashboard, overview view"

> "Check team five-element energy balance"

服务器状态

Port: 3000. MCP endpoint: http://localhost:3000/mcp/manifest. Health: http://localhost:3000/health. Data persistence: enabled (JSON files, auto-save/load).

Available tools

listMembers

List all registered team members with their roles and skills.

registerMember

Register a new team member with skill sets and five-element profile.

createTask

Create a new task with description, five-element attributes, priority, and required skills.

matchMembers

Smartly match the best members to unassigned tasks based on skills and attributes.

assignTask

Assign a specific task to a chosen member and update task status.

updateProgress

Update the progress of a task and reflect changes on the team dashboard.

viewDashboard

View a team dashboard that summarizes tasks, members, and progress.

checkBalance

Check the five-element energy balance across the team to ensure balanced collaboration.