MATLAB Slash Commands

Slash commands for MATLAB development workflows across multiple AI coding agents. Provides documentation generation, testing, optimization, code analysis, and live execution via MCP.

Multi-Agent Support

Agent	Format	Commands	Install
Claude Code	Plugin	All 13	/plugin install github:matlab/slash-commands
GitHub Copilot	.prompt.md	6 key commands	Copy copilot/prompts/ to .github/prompts/
Cursor IDE	.mdc rules	Rules (always-on)	Copy cursor/rules/ to .cursor/rules/
Any MCP agent	MCP server	Via MATLAB MCP Core Server	See MCP Guide

MCP Integration

When the MATLAB MCP Core Server is connected, commands gain live capabilities:

MCP Tool	What It Enables
check_matlab_code	Real Code Analyzer diagnostics
evaluate_matlab_code	Execute MATLAB expressions, benchmark code
run_matlab_file	Run .m files and return output
run_matlab_test_file	Execute test suites and report results
detect_matlab_toolboxes	List installed toolboxes

All commands work without MCP. They generate code and instructions as before. MCP adds validation, execution, and measurement. See MCP Integration Guide.

Installation

Claude Code (Full Support)

/plugin install github:matlab/slash-commands

GitHub Copilot

cp -r copilot/prompts/ .github/prompts/

Cursor IDE

mkdir -p .cursor/rules
cp cursor/rules/matlab-development.mdc .cursor/rules/

Manual / Development

git clone https://github.com/matlab/slash-commands.git

Available Commands

Core Commands

Command	Description	MCP Enhanced
/matlab-readme	Generate project README	Yes
/matlab-docs	Document MATLAB functions	Yes
/matlab-livescript	Create formatted Live Scripts	Yes
/matlab-check	Run Code Analyzer on code	MCP
/matlab-run	Execute MATLAB code or files	MCP

Testing Commands

Command	Description	MCP Enhanced
/matlab-test	Generate unit tests	Yes
/matlab-property-test	Create property-based tests	Yes
/matlab-run-tests	Run test suite and report results	MCP

Quality Commands

Command	Description	MCP Enhanced
/matlab-optimize	Optimize code performance	Yes
/matlab-errors	Add error handling and validation	Yes

Development Commands

Command	Description	MCP Enhanced
/matlab-explain	Explain code functionality	Yes
/matlab-refine	Refine requirements into specs	Yes
/matlab-toolboxes	Detect installed toolboxes	MCP

MCP Enhanced = works without MCP but gains live capabilities with it MCP = primarily designed for MCP, with static fallback

Quick Start

Generate and Validate Tests (with MCP)

/matlab-test myFunction        # Generate tests
/matlab-run-tests              # Run them via MCP, see pass/fail

Optimize with Real Benchmarks (with MCP)

/matlab-optimize               # Optimizes and benchmarks before/after

Full Development Pipeline

/matlab-refine "process sensor data"    # Clarify requirements
/matlab-test processData                # Generate tests
/matlab-check processData.m            # Run Code Analyzer
/matlab-optimize                        # Optimize performance
/matlab-errors                          # Add error handling
/matlab-docs processData               # Generate documentation
/matlab-readme                          # Generate project README

Documentation

• Command Reference: Reference for all 13 commands
• MCP Integration Guide: Setup and usage with MATLAB MCP Core Server
• Cross-Agent Support: Using commands with Copilot, Cursor, etc.
• Usage Guide: Development patterns and best practices
• Installation Guide: Setup instructions for all agents
• Examples: Complete workflow examples

Requirements

• An AI coding agent (Claude Code, GitHub Copilot, Cursor, or any MCP-compatible agent)
• MATLAB development environment (for running generated code)
• Optional: MATLAB MCP Core Server for live execution

Contributing

See CONTRIBUTING.md for guidelines on adding commands, improving existing ones, and maintaining cross-agent formats.

License

BSD 3-Clause License. See LICENSE.

Copyright (c) 2025, The MathWorks, Inc.

Related Projects

• matlab/matlab-mcp-core-server: MATLAB MCP server for AI agent integration
• matlab/prompts: MATLAB prompts for AI assistants and copilots
• matlab/skills: Agents Skills for MATLAB