ML Research

Provides data visualization and research productivity tools via MCP for ML researchers and developers.

python

0

GitHub Stars

python

Language

4 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": {
    "nishide-dev-ml-research-mcp": {
      "command": "uvx",
      "args": [
        "--from",
        "git+https://github.com/nishide-dev/ml-research-mcp",
        "ml-research-mcp"
      ],
      "env": {
        "YOUR_API_KEY": "<YOUR_API_KEY>"
      }
    }
  }
}

ML Research MCP is a modular server that powers AI-assisted machine learning research workflows by providing an extensible set of tools for data visualization and future research tasks. It enables you to generate publication-quality plots, perform statistical visualizations, and integrate smoothly with MCP-enabled AI assistants to streamline your scientific explorations.

How to use

You interact with ML Research MCP through an MCP client. Install and run the MCP server, then connect your client to it to request plots and visualizations. Once connected, you can ask the assistant to generate publication-ready figures, such as line plots, heatmaps, and histograms, or to render 2D representations from your data. You can provide data directly or reference files, and you will receive image outputs in your preferred formats (PNG, PDF, SVG) suitable for inclusion in papers and presentations.

How to install

Prerequisites: ensure you have Python 3.12 or newer and a compatible MCP client environment installed on your workstation.

# Quick start using Claude Code style integration
claude mcp add-json "ml-research" \
  '{"command":"uvx","args":["--from","git+https://github.com/nishide-dev/ml-research-mcp","ml-research-mcp"]}'

# Verify installation
claude mcp list

# Direct execution with uvx (alternative approach)
uvx --from "git+https://github.com/nishide-dev/ml-research-mcp" ml-research-mcp

# Or run from a local clone (development)
cd /path/to/ml-research-mcp
uvx --from . ml-research-mcp
"}]} ,{

Configuration and usage notes

The server is designed to be started via the MCP client and supports both remote (HTTP-like) and local stdio configurations. In this setup, you typically run the server using the UVX command with the repository URL and the target MCP name.

Examples and practical workflows

Generate a publication-quality line plot from a CSV file with temperature over time.
Create a heatmap representing a correlation matrix and export as PDF.
Produce a scatter plot with custom x, y, and size mappings for experimental groups.

Available tools

plot_line

Generate a publication-quality line plot from x and y data, with optional styling and output settings.

plot_scatter

Create a scatter plot with optional size and color mappings to represent multi-dimensional data.

plot_bar

Produce categorical bar charts with vertical or horizontal orientation.

plot_histogram

Render histograms with adjustable bins and density options for distribution analysis.

plot_box

Display statistical summaries and identify outliers using box plots.

plot_violin

Show distribution shapes with kernel density estimation alongside box plot features.

plot_heatmap

Visualize matrix data as a heatmap with optional annotations.

plot_contour

Render contour plots to represent 3D data in a 2D plane.

plot_pcolormesh

Provide fast pseudocolor plots for large datasets with shading options.