CRC-LNM Multimodal Research Assistant MCP

This MCP server provides a six-tool, research-assistance workflow for allowlisted, deidentified CRC-LNM cases. It accepts only precomputed 1409-dimensional CT features, 768-dimensional pathology features, and four clinical values. It does not accept raw imaging files, file paths, or external feature vectors.

ModelScope Hosted STDIO Configuration

Publish the wheel to PyPI before selecting ModelScope hosted STDIO deployment. The first and only server configuration is valid JSON without comments and uses the PyPI package rather than repository files or local paths.

{
  "mcpServers": {
    "crc-lnm-research-assistant": {
      "command": "uvx",
      "args": [
        "crc-lnm-medical-agent@1.0.5",
        "--transport",
        "stdio"
      ],
      "env": {
        "UV_TORCH_BACKEND": "cpu"
      }
    }
  }
}

When entering this configuration in the ModelScope web form, enter uvx in the command field only. Enter crc-lnm-medical-agent@1.0.5, --transport, and stdio as separate arguments. Do not paste a combined shell command such as uvx crc-lnm-medical-agent@1.0.5 --transport stdio into the command field. Set UV_TORCH_BACKEND=cpu in the server environment so the hosted Linux installation resolves CPU PyTorch packages instead of CUDA runtime packages.

The published wheel contains the immutable model bundle and trusted release JSONL. On first launch it creates a verified case-package cache and transient artifacts in a writable system cache directory. No local path argument is required. Set CRC_LNM_MCP_RUNTIME_ROOT only when an operator needs a different writable cache location.

Related MCP server: DROMA MCP Server

Verification Order

1. Build and inspect the wheel, then run the console entry point from an unrelated working directory.

2. Publish the verified wheel to PyPI and start it with the exact uvx command above.

3. Let ModelScope complete list_tools, then manually test each required tool.

4. Obtain the ModelScope URL, add it as a Nexent custom MCP service, enable the six tools, debug the agent, and verify a post-publication question.

docs/PLATFORM_DEPLOYMENT.md covers the separate authenticated Streamable HTTP container path. 使用说明.md documents the local release workflow and constraints.