BCI-MCP is a Brain-Computer Interface server that enables AI systems to receive and process neural signals through the Model Context Protocol. It supports multiple EEG devices (OpenBCI, Emotiv, NeuroSky), provides real-time signal processing with features like bandpass filtering and event detection, and exposes a WebSocket API for streaming brain activity data. The implementation includes tools for device connection, calibration, data recording, and visualization, making it valuable for creating more intuitive human-AI interactions where brain activity can provide additional context to language models.
No reviews yet. Be the first to review!
Sign in to join the conversation
Highlight AILists all available EEG devices that can be connected to the MCP server.
Establishes a connection to an EEG device.
Disconnects from the currently connected EEG device.
Retrieves the current state of the brain metrics: focus, calm, attention, and signal quality.
Returns the power levels of different EEG frequency bands.
Checks the quality of the EEG signal.
Provides a list of definitions for the metrics used in the system.
Calibrates the EEG device to ensure accurate readings.
Records EEG data for a specified duration.
Begins a neurofeedback session on a specified metric (e.g., focus, calm).
Retrieves the score from the most recent neurofeedback session.
Marks a specific event in the EEG data for later analysis.
Provides a summary of the current streaming state and metrics.
