matlab-agentic-toolkit  by matlab

Agentic toolkit for MATLAB engineering

Created 1 month ago
535 stars

Top 58.7% on SourcePulse

GitHubView on GitHub
Project Summary

Summary

The MATLAB Agentic Toolkit enhances AI coding agents by embedding expert MATLAB knowledge, workflows, and best practices. It solves the problem of agents lacking domain-specific expertise, which leads to inefficient code and missed functionalities. By providing agents with direct MATLAB execution capabilities and curated skills, the toolkit enables them to generate idiomatic code, perform complex tasks, and integrate seamlessly into engineering and scientific workflows, significantly reducing development time and token consumption.

How It Works

The toolkit establishes a live connection to MATLAB for code execution, testing, and result analysis. It supplements this with a curated "skills" catalog, which encapsulates expert MATLAB knowledge and workflows. AI agents consult relevant skills and then utilize the MATLAB connection to execute tasks, ensuring efficient and expert-level operation within the MATLAB environment.

Quick Start & Requirements

  • Install/Run: Clone (git clone https://github.com/matlab/matlab-agentic-toolkit.git) and instruct your agent to "Set up the MATLAB Agentic Toolkit." For Claude Code users without cloning, use claude plugin marketplace add "https://github.com/matlab/matlab-agentic-toolkit" then claude plugin install matlab-core@matlab-agentic-toolkit.
  • Prerequisites: MATLAB R2020b+, supported AI agent (Claude Code, Copilot, Codex, Gemini CLI, Sourcegraph Amp), Git.
  • Docs: Getting Started guide.

Highlighted Details

  • MCP Core Server Tools: evaluate_matlab_code, run_matlab_file, run_matlab_test_file, check_matlab_code, detect_matlab_toolboxes.
  • Expertise: Skills catalog covers Image Processing, App Building, Software Development, Signal Processing, Wireless Communications.
  • Platform Integration: Automated setup for Claude Code, Copilot, Codex, Gemini CLI, Sourcegraph Amp.

Maintenance & Community

Feedback and issues are managed via GitHub Issues. While pull requests are reviewed, external contributions are not merged. Support is available through GitHub issues or MathWorks technical support.

Licensing & Compatibility

The toolkit is governed by the MathWorks Software License Agreement. Use of MCP servers is restricted to MATLAB environments and requires explicit contact with MathWorks for any shared or centralized server configurations.

Limitations & Caveats

The toolkit is under active development; report issues encountered. Security requires careful human oversight of all tool calls. For Simulink integration, a separate toolkit and experimental installer are available. External pull requests are not merged into the repository.

Health Check
Last Commit

6 days ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
9
Star History
405 stars in the last 30 days

Explore Similar Projects

Feedback? Help us improve.