feat: Add MATLAB MCP server

[XiaoliangQi]

User description

This commit introduces a new MCP server, matlab-server, allowing agents to execute MATLAB code.

Key features:

• Executes MATLAB code via matlab -batch.
• Supports both text output and capturing plots as PNG images.
• Includes comprehensive README for setup and usage.
• Provides unit tests with mocking for MATLAB interaction.

The server structure is based on the mathematica-check server. It checks for MATLAB availability and handles errors during execution.

Includes:

• Server logic in servers/matlab-server/src/matlab_server/server.py.
• Project files: pyproject.toml, setup.py.
• Documentation: servers/matlab-server/README.md.
• Tests: servers/matlab-server/tests/test_server.py and pytest.ini.
• Python version file: servers/matlab-server/.python-version (set to 3.11.8).

---

PR Type

Enhancement

---

Description

• Add complete MATLAB MCP server implementation

• Support text output and PNG plot generation

• Include comprehensive test suite with mocking

• Provide detailed documentation and setup instructions

---

Changes walkthrough 📝

	Relevant files
Configuration changes	5 files setup.py Package setup configuration for MATLAB server                        +17/-0    .python-version Update Python version to 3.11.8                                                    +1/-1      .python-version Set Python version for MATLAB server                                          +1/-0      pyproject.toml Project metadata and dependencies configuration                    +16/-0    pytest.ini Pytest configuration for testing                                                  +3/-0
Miscellaneous	1 files __init__.py Python package initialization file                                              +2/-0
Enhancement	1 files server.py Core MATLAB MCP server implementation                                        +277/-0
Tests	1 files test_server.py Comprehensive test suite with mocking                                        +316/-0
Documentation	1 files README.md Complete documentation and setup guide                                      +154/-0

---

Need help?

Type /help how to ... in the comments thread for any questions about Qodo Merge usage.

Check out the documentation for more information.

[qodo-merge-pro]

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 4 🔵🔵🔵🔵⚪

🧪 PR contains tests

🔒 Security concerns Command injection: The code parameter from user input is directly embedded into MATLAB command strings without proper sanitization or escaping. An attacker could potentially inject malicious commands by crafting MATLAB code that breaks out of the intended context. For example, malicious code could include MATLAB commands that interact with the file system or execute system commands through MATLAB's system() function. The code should implement proper input validation and consider using safer execution methods.

⚡ Recommended focus areas for review Security Risk The code directly passes user input to subprocess execution without proper sanitization. The code parameter is inserted into MATLAB command strings that are executed via subprocess, potentially allowing command injection attacks. matlab_command_code = f""" try {code} % Check if a figure exists before trying to save if ~isempty(get(groot,'CurrentFigure')) print('-dpng', '{safe_plot_file_path}'); else disp('MATLAB_WARNING: No figure was generated to save.'); end catch ME disp(['MATLAB_ERROR: ' ME.message]); % Optional: rethrow(ME); % if we want MATLAB to exit with error code for script errors end exit; % Ensure MATLAB exits """ Resource Leak Temporary files created by tempfile.mkstemp() may not be properly cleaned up if exceptions occur before reaching the finally block. The file descriptor is closed but the file cleanup in finally block might not execute in all error scenarios. fd, temp_plot_file_path = tempfile.mkstemp(suffix=".png") os.close(fd) # Close the file descriptor, MATLAB will open and write to it. temp_plot_file = temp_plot_file_path Error Handling The global state variables for MATLAB availability checking are not thread-safe and could lead to race conditions in concurrent environments. Multiple simultaneous calls could interfere with each other's state. _matlab_checked = False _matlab_available = False async def check_matlab_installation() -> bool: """Checks if MATLAB is installed and accessible.""" global _matlab_checked, _matlab_available if _matlab_checked: return _matlab_available

[Copilot]

Pull Request Overview

This PR introduces a new MATLAB MCP server to execute MATLAB code and capture both text and plot outputs. Key changes include adding the server logic and MCP tool in Python, a comprehensive suite of unit tests using async mocks, and updated configuration files (setup.py, pyproject.toml, pytest.ini, and README).

Reviewed Changes

Copilot reviewed 9 out of 10 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
servers/matlab-server/tests/test_server.py	Comprehensive unit tests covering MATLAB installation checks, text/plot code execution, and global state resets.
servers/matlab-server/src/matlab_server/server.py	Server logic for checking MATLAB availability, executing MATLAB code, and handling plot output.
servers/matlab-server/src/matlab_server/init.py	Package initialization file.
servers/matlab-server/setup.py	Build and packaging configuration for the MATLAB server.
servers/matlab-server/pytest.ini	Pytest configuration ensuring proper module discovery and asyncio mode setup.
servers/matlab-server/pyproject.toml	Project metadata and dependency configuration.
servers/matlab-server/README.md	Updated documentation detailing installation, usage, and integration with MCP clients.
servers/matlab-server/.python-version	Specifies the required Python version (3.11.8) for the MATLAB server.
.python-version	Updated global Python version to 3.11.8.

[Copilot]

Consider using textwrap.dedent to build multiline MATLAB command strings so that unintended leading whitespace does not affect MATLAB's interpretation of the script.

Suggested change

	matlab_command_code = f"""
	matlab_command_code = textwrap.dedent(f"""

[qodo-merge-pro]

PR Code Suggestions ✨

Explore these optional code suggestions:

Category	Suggestion	Impact
Security	Fix temporary file race condition Closing the file descriptor immediately after mkstemp() creates a race condition where another process could potentially claim the file path before MATLAB writes to it. Consider using tempfile.NamedTemporaryFile with delete=False for safer temporary file handling. servers/matlab-server/src/matlab_server/server.py [106-107] -fd, temp_plot_file_path = tempfile.mkstemp(suffix=".png") -os.close(fd) # Close the file descriptor, MATLAB will open and write to it. +temp_file = tempfile.NamedTemporaryFile(suffix=".png", delete=False) +temp_plot_file_path = temp_file.name +temp_file.close() # Close but don't delete Apply / Chat Suggestion importance[1-10]: 9 __ Why: The suggestion correctly identifies a potential race condition (TOCTOU vulnerability) when creating a temporary file. The proposed solution using tempfile.NamedTemporaryFile with delete=False is a much safer and more robust way to handle this, making it a critical security improvement.	High
General	Handle critical stderr messages properly The code continues execution after logging a stderr warning, but stderr from MATLAB often indicates serious issues that should be handled more carefully. Consider evaluating if the stderr content warrants stopping execution or returning an error. servers/matlab-server/src/matlab_server/server.py [174-178] if stderr_str: # Process-level stderr logger.warning(f"MATLAB process produced stderr: {stderr_str}") - + # Consider if stderr indicates a critical error that should halt execution + if any(critical_term in stderr_str.lower() for critical_term in ['error', 'failed', 'exception']): + raise RuntimeError(f"MATLAB execution failed with critical stderr: {stderr_str}") if output_format == "plot": Apply / Chat Suggestion importance[1-10]: 7 __ Why: The suggestion correctly points out that stderr from the MATLAB process might contain critical errors that should halt execution. While the current code logs stderr as a warning, explicitly checking for error keywords and raising a RuntimeError would make the error handling more robust and prevent silent failures.	Medium
More