wegene-assistant MCP server

MCP server for WeGene Assistant, using LLM to analyze a user’s WeGene genetic testing report.

Components

Resources

Once a user is authorized, all the reports under his/her account will be exposed as a resource:

• Custom wegene:// URI scheme for accessing each individual report
• A report resource has a name, description and application/json mimetype

Tools

The server implements one tool:

• wegene-oauth: Start a WeGene Open API oAuth process in the browser
  • The user should complete the authorization in 120 seconds so LLM will be able to further access the reports.
• wegene-get-profiles: Read the profile list under a user’s WeGene account
  • Profiles’ name and id will be returned for LLM to use.
• wegene-get-report-info: Return the report meta info so LLM will know what reports are available.
  • A list of report names, descriptions, endpoints, etc. will be returned
• wegene-get-report: Read the results of a single report under a profile
  • Returns the result JSON specified in WeGene’s Open API platform
  • Arguements
    • report_endpoint: The report’s endpoint to be retrieved from
    • report_id: The report’s id to be retrieved
    • profile_id: The profile id to retrieve report from

Configuration

• You will need WeGene Open API key/secret to use this project.
• Copy .env.example as .env and update the key and secret in the file.

Quickstart

Install

Installing via Smithery

To install WeGene Assistant for Claude Desktop automatically via Smithery:

npx -y @smithery/cli install @xraywu/mcp-wegene-assistant --client claude

Insall Locally

Prepare MCP Server

1. Clone this project
2. Run uv sync --dev --all-extras under the project’s root folder

Claude Desktop Configuration

• On MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• On Windows: %APPDATA%/Claude/claude_desktop_config.json

Add below contents in the configuration file:

{
  "mcpServers": {
    "wegene-assistant": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/wegene-assistant",
        "run",
        "wegene-assistant"
      ]
    }
  }
}