Quantdinger

Run the install script, which clones the repo to ~/quantdinger and pulls all Docker images. Default credentials are quantdinger / 123456; change them immediately after first login.

curl -fsSL https://raw.githubusercontent.com/brokermr810/QuantDinger/main/install.sh | bash

Copy the example env file to .env and open it. You must change SECRET_KEY (the service refuses to start if it matches the example value). Add your exchange API keys and broker credentials as needed.

cd ~/quantdinger/backend_api_python
cp env.example .env
./scripts/generate-secret-key.sh   # outputs a new SECRET_KEY value to paste into .env

Launch all services with Docker Compose. The web UI will be available at http://localhost:8888 within about 60 seconds.

cd ~/quantdinger
docker compose up -d

Install quantdinger-mcp from PyPI into your local Python environment. This package registers the MCP server that your AI client will connect to.

pip install quantdinger-mcp

Log in to the web UI at http://localhost:8888, navigate to Settings > Agent Tokens, and create a new token. By default, tokens are restricted to paper trading. To enable live execution you must also set AGENT_LIVE_TRADING_ENABLED=true in your .env.

Add the QuantDinger MCP server to your Claude Desktop or Cursor config. Replace the placeholder values with your actual local URL and agent token.

{
  "mcpServers": {
    "quantdinger": {
      "command": "python",
      "args": ["-m", "quantdinger_mcp"],
      "env": {
        "QUANTDINGER_URL": "http://localhost:8888",
        "QUANTDINGER_AGENT_TOKEN": "your-agent-token-here"
      }
    }
  }
}