Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This collection provides a structured set of capabilities for engineers to coordinate complex scientific computations. It enables the automation of aerospace environment modeling and coordinate conversions, alongside tools for statistical classification and reliability analysis.
By integrating these specific mathematical functions into an agentic framework, the toolkit helps aerospace engineers and data scientists move from raw tabular data to lifetime predictions and environmental simulations with greater precision.
WHO IT'S FOR
Aerospace engineers
automate aerospace environment and coordinate computations
Reliability engineers
perform reliability analysis and lifetime prediction
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Follow the documented setup, then try a first task.
I’m using
MATLAB Agentic Toolkit Installer
Installs: Installs the MATLAB MCP Server and adds whichever MATLAB Agentic Toolkit skill groups the user selects in the installer. · Claude Code, GitHub Copilot, OpenAI Codex, Gemini CLI, Amp
Before you start
MATLAB R2021a or later
An AI coding agent that supports MCP servers and skills
Download agenticToolkitInstaller.mltbx using the repository's release link.
Open the downloaded agenticToolkitInstaller.mltbx file with MATLAB to install the installer add-on.
Enter this command in MATLAB.
setupAgenticToolkit("install")
In the installer, select only the skill groups relevant to your work. Re-run the installer later if you need additional groups.
Optionally, to connect the agent to the currently open MATLAB session instead of creating a new session, enter this command in the MATLAB command window. With multiple sessions, run it in the session you want the agent to use.
shareMATLABSession()
To verify installation, ask your agent: "What version of MATLAB is running? List the installed toolboxes."
Replace <path-to-code> with the MATLAB function or class to test.
Provide expected behavior and known error conditions if they are not evident from the code.
Identify external dependencies or reference data that may require fixtures or mocks.
Check the repository for this skill’s setup.
Use the matlab-write-tests skill to inspect the MATLAB function or class at <path-to-code>. Before writing any test code, propose a concise class-based test plan covering its expected behavior, important edge cases, error conditions, and required fixtures or mocks. Wait for my approval after presenting the plan.
Replace <path-to-workload> with a MATLAB script or function that uses parfor, parfeval, or spmd.
MATLAB R2020a or later.
A local machine rather than a cluster; this workflow is not for parsim, distributed arrays, or codistributed arrays.
Check the repository for this skill’s setup.
Use the matlab-use-thread-pool skill with the local MATLAB workload at <path-to-workload>. Check the current pool, switch to a thread-based pool only if necessary, run the existing parfor, parfeval, or spmd workload without refactoring it, and report whether it succeeds or the exact MATLAB diagnostic if it fails.
MATLAB R2024b or later for the documented validateGPU workflow.
Access to the MATLAB installation and host whose GPU availability should be checked.
Check the repository for this skill’s setup.
Use the matlab-setup-gpu skill to check whether this MATLAB installation can use a GPU. Run canUseGPU first. If it returns false, run validateGPU("all") and summarize the reported failure; if it returns true, run gpuDeviceTable and summarize the visible devices without selecting a different GPU.
Claude
Codex
Gemini CLI
Copilot
Amp