MLflow MCP
JSON →MLflow MCP server for ML experiment tracking with advanced querying, run comparison, artifact access, and model registry.
Install
npx @modelcontextprotocol/inspector Tools · 13
- search_experiments Search for MLflow experiments by name or tags
- get_experiment Get details of a specific MLflow experiment
- create_experiment Create a new MLflow experiment
- search_runs Search for runs in an MLflow experiment
- get_run Get details of a specific MLflow run
- log_param Log a parameter to an MLflow run
- log_metric Log a metric to an MLflow run
- log_artifact Log an artifact to an MLflow run
- list_artifacts List artifacts in an MLflow run
- download_artifact Download an artifact from an MLflow run
- search_model_registry Search for registered models in MLflow Model Registry
- get_model_version Get details of a specific model version
- transition_model_version_stage Transition a model version to a new stage
Environment variables
MLFLOW_TRACKING_TOKEN
Links
★ 12 GitHub stars