MathModel Skill Standard
Ten skills for Codex, Claude Code and Trae carry a mathematical modeling contest paper from the problem statement to a PDF check.
- 458 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The agent runs calculation code and creates files in the project directory
- The result does not replace human review, and the participant checks contest AI rules personally
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/mathmodel-skill-standardDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent runs calculation code and creates files in the project directory
- The result does not replace human review, and the participant checks contest AI rules personally
This is third-party code. Review the repository files before installing.
What it does
A package of ten skills and one entry router drives the work through stages S0-S8: pre-checking the problem and attachments, analyzing the statement, choosing a model, preparing data and charts, running real code, checking evidence, writing section by section and verifying Word and PDF output. Stages hand files to one another with SHA-256 checksums. The author stresses that this is not a one-step paper prompt: the agent must run real code and save results, and writing begins only after the checks pass. The package includes `install.py`, which copies the skills into `.agents/skills/` for Codex, `.claude/skills/` for Claude Code or `.trae/skills/` for Trae.
Who it is for. Participants in mathematical modeling contests who solve the problem together with an agent, and team leads who need reproducible calculations and checks before submission.
Good fit when
- A task with spreadsheet attachments has to be taken by an agent to a finished paper with calculations and charts.
- An interrupted run has to resume from the stopping point without losing evidence.
- Before submission every chart and conclusion must be backed by code that was actually run.
- A single chapter has to be fixed without rebuilding the whole text.
Not a fit when
- A quick draft of a few pages is wanted: the author has the Lite and Flash branches for that.
- Contest rules forbid AI help: compliance rests on the participant.
- The task does not involve modeling with computation: the whole chain is built around running code.
Example request
Start on the task in `problem_files/` with the orchestrator: run the pre-check, propose a model route and stop before running code.Limitations
Version 2.4.0-dev.4 is called a development branch in the README, with no released build. Python 3.11 or 3.12 is needed, plus LibreOffice for the final PDF check. The installer does not overwrite old installs and asks you to sort out conflicts by hand. Model recommendations are not backed by full runs, as the README says. Evidence checks do not guarantee a high score or a win. Descriptions, examples and instructions are mostly in Chinese. Only one edition is installed at a time: Standard, Lite, Flash or Pro.
How to disable. Delete the installed skills from the project's `.agents/skills/`, `.claude/skills/` or `.trae/skills/` and the `.mathmodel/install.json` file. Result folders and `problem_files/` stay and are removed by hand.
FAQ
Which editions exist and how do they differ?
Flash gives a quick basic draft. Lite gives a basic report with minimal steps. Standard drives a full paper with Word and PDF checks, and Pro adds repeated verification and several routes. Each edition lives in its own Git branch.
What does `install.py` do?
It copies the skills into the selected platform's directory, records file hashes and never touches the network. With `--check` it only checks whether installation is possible.
What else is needed besides Python?
LibreOffice for the final PDF check. Dependencies come from `requirements.txt`, and for optimization, if needed, from `requirements-modeling.txt`.
Can Standard and Lite be installed together?
No. The README requires one edition and one install location per project, otherwise a mixed install appears.
Does the package guarantee an award?
No. Passing the checks does not equal high quality or a win, as the README says outright.
Discussion
No questions or reviews yet.
Related
SkillsA skill for math modeling contests CUMCM, MCM/ICM and the Electrical Cup: 10 stages, three roles, five gates and a LaTeX paper
Two skills for the CUMCM and MCM/ICM math modeling contests: problem breakdown, model choice, code and the paper
A workflow for mathematical modeling contests: ten stages, ten gates and frozen results before the paper is written.
A three-stage workflow for math modeling tasks: analysis and model choice, Python or MATLAB computation, and paper assembly in Word or LaTeX
A skill set for math modeling contests: problem parsing, method selection, Python and MATLAB code, and paper writing
Eight skills for math modeling contests that carry a task from the problem to a paper with figures and checked references
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| MathModel Skill Standard (this card) | no | 458 | |
| MathModeling-skills | no | 1,091 | mathmodeling-skills bundles analysis, methods and formatting for Python and MATLAB; this one has stages S0-S8 with evidence gates and Word and PDF checks |
| mathmodel-skill | no | 298 | mathmodel-skill by another author leads a team through CUMCM, MCM/ICM and an electrical contest in ten stages; this is a separate package with an installer and Word and PDF checks |
README in short
The README opens with a table of editions and model recommendations, then covers installation and project layout. Detailed writing rules, stage contracts and output layout live in `docs/`.
README badge
Original SKILL.md
--- name: MathModel Skill Standard description: "数学建模全流程技能包：赛题与附件预检、题意与模型路线、数据与图表计划、真实运行建模代码、证据门禁、分章节正式写作与 Word/PDF 渲染检查，产出可复现的竞赛论文和全套证据。当用户要做完整篇数学建模竞赛论文时使用。" license: MIT compatibility: "Python 3.11-3.12 与 requirements.txt；最终 PDF 渲染检查需要 LibreOffice；优化求解可选 scipy/HiGHS。" ---