Menu
Use when a computational math repository or task uses MATLAB files, MATLAB README commands, MATLAB toolboxes, or MATLAB MCP execution tools.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Route AI4Math computational mathematics tasks to the appropriate bundled skill package.
Use when reproducing, deploying, diagnosing, tuning, visualizing, or reporting computational math research code with a coding agent
This skill should be used when the user asks to search for external algorithms, find related implementations, discover alternative approaches, or look up papers and code for computational math optimization methods.
This skill should be used when the user asks to tune parameters, run hyperparameter search, optimize algorithm settings, or find the best configuration for a computational math method.
Use when classifying a computational math repository or task into a broad domain before choosing specialist Skills, references, runtimes, and validation evidence.
Use when a user starts an end-to-end computational math research-code reproduction, deployment, tuning, visualization, or reporting workflow with a coding agent.
| name | matlab-runtime-skill |
| description | Use when a computational math repository or task uses MATLAB files, MATLAB README commands, MATLAB toolboxes, or MATLAB MCP execution tools. |
MATLAB is a runtime backend, not the workflow driver. The workflow remains Skill-first and conversation-first: computational_math_reproduction_workflow_skill writes plans, records evidence under outputs/{run_id}/, and pauses for human approval before consequential execution.
This Skill defines how to inspect, plan, and optionally execute MATLAB code in computational math reproduction tasks. The local helper scripts/matlab_runtime.py can summarize MATLAB files, README commands, likely toolboxes, entrypoint candidates, and local matlab/octave executable availability.
.m, .mlx, .mat, MATLAB project files, or MATLAB README commands.matlab_environment_setup_skill first.matlab or octave is available, generate a CLI run plan and ask for run-plan approval before execution.outputs/{run_id}/logs/run.log.| Capability | Use |
|---|---|
detect_matlab_toolboxes | Confirm MATLAB version and installed toolboxes. |
check_matlab_code | Static Code Analyzer checks for .m files. |
run_matlab_file | Run scripts or programs from files. Prefer this over long inline code. |
run_matlab_test_file | Run MATLAB unit tests with structured results. |
evaluate_matlab_code | Short diagnostics and quick variable/toolbox checks only. |
Use this helper for static preflight and plan evidence:
conda run -n ai4math python -m skills.matlab_runtime_skill.scripts.matlab_runtime --source <repo> --out outputs/{run_id}
The helper writes matlab_runtime_summary.json when --out is provided. It does not execute MATLAB code. repo_analyzer.py and run_planner.py use the same helper logic to attach MATLAB runtime summaries and generate approved matlab -batch or octave --eval candidate plans when a local executable is available.
references/INDEX.md.matlab, or local octave are available.matlab_environment_setup_skill for agent-neutral setup.plan.md and wait for approval.failure_diagnosis_skill.The official MATLAB Agentic Toolkit recommends registering MATLAB MCP with Codex using:
codex mcp add matlab -- "<MCP_SERVER_PATH>" --matlab-root "<MATLAB_ROOT>" --matlab-display-mode "<DISPLAY_MODE>"
After registration, the MATLAB MCP server should have a longer timeout such as tool_timeout_sec = 600 in the Codex config. Treat setup as a user-approved environment change, not an automatic reproduction step.