No EEG hardware needed — the synthetic device generates realistic brainwaves so you can run the full demo with just pip install.

1. Install¶

git clone https://github.com/enkhbold470/bci-mcp.git
cd bci-mcp
pip install -e ".[all]"

2. Check available devices¶

bci-mcp devices

Lists every URI scheme registered (synthetic, neurofocus, brainflow, lsl, serial, playback) and any auto-discovered hardware.

3. Live terminal brain-meter¶

bci-mcp stream --device synthetic://

Press Ctrl-C to stop. Add --once to print a single snapshot:

bci-mcp stream --device synthetic:// --once

4. Record a session¶

bci-mcp record --device synthetic:// --seconds 30 --out session.npz

Formats: .npz (default), .csv, .edf (needs [edf] extra). Replay with:

bci-mcp play session.npz

5. Neurofeedback trainer¶

bci-mcp neurofeedback --device synthetic:// --metric focus --target 0.7 --seconds 60

Prints a live indicator showing whether your metric is in-zone, then a session summary.

6. Web dashboard¶

bci-mcp dashboard

Opens at http://127.0.0.1:8000 — shows live band powers and metrics.

7. MCP server for Claude Desktop¶

bci-mcp serve

Or point Claude Desktop at it in claude_desktop_config.json:

{
  "mcpServers": {
    "bci-mcp": {
      "command": "bci-mcp",
      "args": ["serve"]
    }
  }
}

Then ask Claude: "Connect to the synthetic brain and tell me my current focus level."

Available CLI commands¶