Scikit-learn

Provides a standardized MCP interface for training, evaluating, and managing Scikit-learn models and datasets.

python

12

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "shibuiwilliam-mcp-server-scikit-learn": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/mcp-server-scikit-learn",
        "run",
        "mcp-server-scikit-learn"
      ]
    }
  }
}

You can run and integrate the Scikit-learn MCP server to train, evaluate, and manage models and datasets through a consistent MCP interface. This server exposes standardized actions for model training, evaluation, preprocessing, persistence, feature engineering, and hyperparameter tuning, helping you build repeatable pipelines and reuse models across projects.

How to use

You connect to the Scikit-learn MCP server using a compatible MCP client. Once connected, you can train models, evaluate performance, prepare datasets, apply preprocessing steps, persist trained models for later reuse, and perform feature engineering or selection as part of an end-to-end workflow. The server exposes common actions that you can chain together to prototype and productionize ML pipelines, check metrics, and tune hyperparameters through cross-validation.

How to install

Prerequisites you need before starting:

Python 3.8+
python -m venv .venv
source .venv/bin/activate # On Windows: .venv\Scripts\activate
pip install -e ".[dev]"
pytest -s -v tests/

Additional notes

To use the server with MCP tooling, install the inspector and register the server as an MCP endpoint according to the tooling instructions. The inspector helps you visualize and debug the MCP interactions during development.

Available tools

train_model

Train a Scikit-learn model using provided training data, parameters, and a specified pipeline.

evaluate_model

Evaluate a trained model against a dataset and return metrics such as accuracy, F1, ROC-AUC, etc.

preprocess_data

Apply preprocessing steps (scaling, encoding, imputation) to input data before training or evaluation.

persist_model

Save a trained model to disk or a model registry for later retrieval.

load_model

Load a previously saved model from storage for inference or evaluation.

feature_engineering

Perform feature engineering and selection to improve model performance.

cross_validation

Run cross-validation to estimate model performance across folds.

hyperparameter_tuning

Tune model hyperparameters automatically using a defined search strategy.