Command Reference

March 25, 2026 · View on GitHub

Reference for all MATLAB Slash Commands, including parameters, options, and examples.

Core Documentation Commands

`/matlab-readme`

Generate project documentation by analyzing the codebase.

Parameters: None

Analyzes:

Project structure and files
Main functions and entry points
Dependencies and requirements
Existing documentation

Generates:

Project overview
Installation instructions
Usage examples
API documentation
Testing instructions

Example:

/matlab-readme

Output: Complete README.md file ready for your repository

`/matlab-docs`

Create MATLAB function documentation following MathWorks standards.

Parameters:

function_name (optional) - Name of function to document

Features:

H1 line and help text
Input/output specifications
Usage examples
Error conditions
See also references

Example:

/matlab-docs calculateMetrics

Output: Complete help text block for the function

`/matlab-livescript`

Generate MATLAB Live Scripts for tutorials and documentation.

Parameters:

topic (optional) - Topic or title for the Live Script

Creates:

Formatted sections with %% markers
Text explanations with % comments
Executable code blocks
Visualization examples
Exercises and solutions

Example:

/matlab-livescript "Signal Processing Tutorial"

Output: Plain text .m file formatted for Live Editor

Testing Commands

`/matlab-test`

Generate unit tests using MATLAB Unit Testing Framework.

Parameters:

function_name (optional) - Function to test

Generates:

Complete test class
Setup/teardown methods
Parameterized tests
Edge case validation
Performance benchmarks

Example:

/matlab-test processData

Output: TestClassName.m file with complete test suite

Test Categories:

Functionality tests
Edge cases
Error conditions
Performance tests
Regression tests

`/matlab-property-test`

Create property-based tests for mathematical invariants.

Parameters:

function_name (optional) - Function to test

Tests:

Mathematical properties
Invariants
Algebraic relationships
Domain constraints

Example:

/matlab-property-test matrixOperation

Properties Tested:

Commutativity
Associativity
Distributivity
Identity elements
Inverse operations
Idempotence

Quality Commands

`/matlab-optimize`

Optimize MATLAB code for performance and efficiency.

Parameters: None (operates on selected code)

Optimizations:

Loop vectorization
Array preallocation
Built-in function usage
Memory optimization
Parallel computing suggestions

Example:

% Select code to optimize, then:
/matlab-optimize

Improvements:

Execution speed
Memory usage
Code readability
Numerical stability

`/matlab-errors`

Add error handling and input validation.

Parameters: None (operates on selected code)

Adds:

Input validation (arguments block)
Try-catch patterns
Custom error messages
Warning management
Resource cleanup

Example:

% Select function to enhance, then:
/matlab-errors

Features:

Type checking
Range validation
Dimension verification
Error recovery
Progress reporting

Development Commands

`/matlab-explain`

Explain MATLAB code functionality and algorithms.

Parameters: None (operates on selected code)

Provides:

Algorithm explanation
Line-by-line breakdown
MATLAB idiom clarification
Performance analysis
Mathematical background

Example:

% Select complex code, then:
/matlab-explain

Explains:

What the code does
How it works
Why design choices were made
Potential improvements

`/matlab-refine`

Transform vague requirements into detailed specifications.

Parameters:

initial_prompt (optional) - Initial requirement description

Generates:

Detailed specifications
Clarifying questions
Implementation checklist
Design decisions
Test scenarios

Example:

/matlab-refine "process sensor data and detect anomalies"

Output:

Complete function specification
Input/output definitions
Performance requirements
Error handling strategy

MCP-Powered Commands

These commands are designed primarily for use with the MATLAB MCP Core Server. They provide static fallback when MCP is not available.

`/matlab-check`

Run MATLAB Code Analyzer on files or code.

Parameters:

file_or_code (optional) - File path or code to analyze

MCP Tools Used: check_matlab_code, evaluate_matlab_code

Features:

Real Code Analyzer diagnostics (with MCP)
Issue categorization: errors, warnings, performance, style
Fix suggestions with corrected code
Static pattern analysis (without MCP)

Example:

/matlab-check myFunction.m

`/matlab-run`

Execute MATLAB code or files via MCP.

Parameters:

file_or_expression (optional) - File path or MATLAB expression

MCP Tools Used: run_matlab_file, evaluate_matlab_code, check_matlab_code, detect_matlab_toolboxes

Features:

Live code execution with formatted results
Pre-execution validation
Error handling with fix suggestions
Execution timing

Example:

/matlab-run "x = magic(5); disp(x)"

`/matlab-run-tests`

Run MATLAB test suite and report results.

Parameters:

test_target (optional) - Test file, folder, or test name

MCP Tools Used: run_matlab_test_file, evaluate_matlab_code, check_matlab_code

Features:

Automated test discovery and execution
Pass/fail/incomplete reporting
Failure analysis with root cause suggestions
Performance reporting

Example:

/matlab-run-tests tests/

`/matlab-toolboxes`

Detect installed MATLAB toolboxes and provide recommendations.

Parameters: None

MCP Tools Used: detect_matlab_toolboxes, evaluate_matlab_code, check_matlab_code

Features:

Installed toolbox inventory
Project dependency analysis
Usage recommendations for unused toolboxes
Missing dependency warnings

Example:

/matlab-toolboxes

Command Options and Flags

While commands don't have explicit flags, they adapt based on context:

Context Detection

Commands automatically detect:

Selected code vs. file context
Function vs. script context
Available MATLAB version
Existing code patterns

Output Customization

Commands generate output considering:

Code style in the project
Existing naming conventions
Documentation patterns
Test framework usage

Output Formats

Code Output

Generated code follows MATLAB conventions:

function output = functionName(input, options)
    arguments
        input (:,:) double
        options.Method string = "default"
    end

    % Implementation
end

Documentation Output

Documentation uses standard MATLAB format:

%FUNCTIONNAME Brief description
%   Detailed explanation...
%
%   See also: RELATED_FUNCTION

Test Output

Tests follow MATLAB Unit Testing Framework:

classdef TestClass < matlab.unittest.TestCase
    methods (Test)
        function testMethod(testCase)
            % Test implementation
        end
    end
end

Error Messages

Commands provide clear error messages:

Error	Meaning	Solution
"No code selected"	Command needs code selection	Select code in editor
"Function not found"	Named function doesn't exist	Check function name
"Invalid MATLAB syntax"	Code has syntax errors	Fix syntax first
"Missing dependencies"	Required toolbox not found	Note requirements

Performance Considerations

Command Execution

Most commands execute instantly
Large codebases may take a few seconds
Commands are stateless (can be rerun)

Generated Code Performance

Commands optimize for:

Execution speed
Memory efficiency
Numerical stability
Readability balance

Version Compatibility

MATLAB Versions

Minimum: R2019b (for arguments blocks)
Recommended: R2021a or later
Legacy support: Available with modifications

Feature Availability

Feature	Required Version
arguments block	R2019b+
string arrays	R2016b+
Live Scripts	R2016a+
Unit Testing	R2013a+

Integration with Other Tools

MATLAB Tools

Commands work with:

MATLAB Editor
Live Editor
Code Analyzer (mlint)
Profiler
Test Runner

Version Control

Generated code is:

Git-friendly (consistent formatting)
Diff-friendly (logical organization)
Merge-friendly (minimal conflicts)

MATLAB MCP Core Server

Commands can use the MCP Core Server for:

Live code execution and validation
Real Code Analyzer diagnostics
Automated test execution
Toolbox detection

See MCP Integration Guide for setup.

Customization

Modifying Commands

Edit command file in commands/ directory
Update frontmatter if needed
Reload plugins in Claude Code

Creating Variations

Copy and modify existing commands:

cp commands/testing/matlab-test.md commands/testing/matlab-integration-test.md
# Edit the new file
# Update plugin.json

FAQ

Multiple files. Commands work on single files or selections. For multiple files, run the command once per file.

MATLAB not installed. Commands generate code without requiring MATLAB. You need MATLAB only to run the generated code.

Modifying existing files. Commands generate new code. You manually integrate it with existing files.

Updating command output style. Edit the command markdown file to customize output format.

Getting Help

Command not working: Check Troubleshooting
Report issues: GitHub Issues
Request features: Discussions