Mne Mcp

Please install the `mne-mcp` MCP server into my current AI client (that's you).

Required prerequisites (do these first if not already done):
- **Install Python 3.10+** — Python 3.10 or newer is required to run the server. (https://www.python.org/downloads/)
- **Install uv, pipx, or another way to run the PyPI package** — The README recommends the portable PyPI-based path. Use uv/uvx for on-demand execution, or pipx/uv tool install for a persistent install. (https://docs.astral.sh/uv/)

Optional prerequisites:
- Optional persistent install for faster startup — Because MNE pulls in a large scientific stack, a persistent install is usually faster than resolving dependencies on each run. Install the package with the ICA extra so ICA tools work out of the box. Run: `pipx install "mne-mcp[ica]"`

Canonical MCP server config (stdio transport):
- command: `uvx`
- args: ["--from","mne-mcp[ica]","mne-mcp","serve","--transport","stdio"]
- optional environment variables:
  - `MNE_MCP_TIMEOUT`: Per-operation timeout in seconds; increase for ICA, time-frequency analysis, or large files. (example: `<timeout-seconds>`)
  - `MNE_MCP_RESULTS_DIR`: Directory where figures and exported objects are saved. (example: `<path-to-results-directory>`)
  - `MNE_MCP_DATA_DIR`: Default directory scanned by mne_list_files. (example: `<path-to-data-directory>`)
  - `MNE_MCP_CONFIG`: Override path to the MNE-MCP config file. (example: `<path-to-config-json>`)

Note: Preferred portable config uses uvx with the PyPI package. If you install persistently via pipx or uv tool install, use command `mne-mcp` with args `["serve", "--transport", "stdio"]` instead. The README also mentions `mne-mcp[full]` as an alternative to enable advanced tools beyond the default ICA extra.

Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.