MCP Server MCP

io.github.us-all/mlflow

By github 👁 1 views ▲ 0 votes

MLflow MCP — experiments, runs, registered models, versions, traces, assessments (MLflow 3)

Homepage Verified source GitHub

Install

npx -y @us-all/[email protected]

Configuration Example

{
  "remotes": [],
  "packages": [
    {
      "registryType": "npm",
      "identifier": "@us-all/mlflow-mcp",
      "version": "1.12.7",
      "transport": {
        "type": "stdio"
      },
      "environmentVariables": [
        {
          "description": "MLflow tracking server URI (e.g. http://localhost:5050, https://mlflow.example.com).",
          "isRequired": true,
          "name": "MLFLOW_TRACKING_URI"
        },
        {
          "description": "Bearer token for authenticated MLflow servers (e.g. Databricks).",
          "isSecret": true,
          "name": "MLFLOW_TRACKING_TOKEN"
        },
        {
          "description": "Basic-auth username (alternative to token).",
          "name": "MLFLOW_TRACKING_USERNAME"
        },
        {
          "description": "Basic-auth password (alternative to token).",
          "isSecret": true,
          "name": "MLFLOW_TRACKING_PASSWORD"
        },
        {
          "description": "Default experiment ID for run-creation tools.",
          "name": "MLFLOW_EXPERIMENT_ID"
        },
        {
          "description": "Comma-separated category allowlist. Default: all categories enabled.",
          "name": "MLFLOW_TOOLS"
        },
        {
          "description": "Comma-separated category disablelist.",
          "name": "MLFLOW_DISABLE"
        },
        {
          "description": "Set to 'true' to enable write/destructive tools. Default read-only.",
          "default": "false",
          "name": "MLFLOW_ALLOW_WRITE"
        }
      ]
    }
  ]
}

mcp model-context-protocol npm