Math-model-skills
Meta-model-agent takes a mathematical modeling task through seven stages: data, model, computation, figures and paper.
- 112 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill runs local scripts and writes files into the working folder.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/math-model-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill runs local scripts and writes files into the working folder.
This is third-party code. Review the repository files before installing.
What it does
Meta-model-agent, a skill for Codex and other agents, builds a workflow for mathematical modeling contests: CUMCM, 51MCM, MCM and ICM. Seven stages run from reading the problem and auditing data to preparing a paper in LaTeX or DOCX and a pre-submission check. Numbers in the text are tied to computation results, and gates close each transition between stages. State saving, rework returns and an audit of the final PDF cover formulas, section references, empty space and small fonts. Instructions live in `SKILL.md` and `references/`.
Who it is for. Students and coaches preparing for mathematical modeling contests, and researchers with similarly structured tasks.
Good fit when
- A team received the problem and attachments and needs a staged modeling plan.
- Numbers in the paper do not match program results and must be tied to computations.
- A paper must be assembled to the contest template with a layout check.
- Interrupted work must resume from the last stage.
Not a fit when
- Contest rules forbid AI help.
- You want a finished paper with no human review.
- The task is far from modeling and computation.
Example request
Begin processing the uploaded CUMCM problem: read the task, audit the data and build a baseline model for the first question.Limitations
The project calls itself a research aid: the author verifies the problem, data, model, programs, references and final materials. Official contest rules outrank repository templates and gates, and whether AI help is allowed must be learned in advance. The primary platform is Codex. Python with dependencies from `scripts/requirements.txt` is needed, and LaTeX for building the paper. MIT license, version v1.3, Codex as the default platform.
How to disable. Remove the skill from the agent global skills directory. Delete working folders with data, programs and the paper created during work separately.
FAQ
Which contests are supported?
CUMCM, 51MCM and MCM/ICM, plus tasks of similar structure. Contest profiles live in `assets/competition_profiles.json`.
How do I install it in Codex?
Ask Codex to add the repository `https://github.com/WuXinbo-bo/Math-model-skills` to global skills. Then upload the problem and data.
Does the skill guarantee a contest result?
No. Internal scores and gates promise nothing about the contest outcome, and the user verifies the materials.
Which paper output formats exist?
PDF and DOCX. LaTeX handles layout, and diagrams can be built through DrawIO and TikZ.
What to do after a failure?
The process supports checkpoint recovery: return to the last saved stage and repeat it, and dependent stages are sent back for rework.
Discussion
No questions or reviews yet.
Related
SkillsA skill for writing a mathematical modeling contest paper: LaTeX template, writing rules and five analyzed winning papers
A skill that reviews a CUMCM competition paper against the problem, scores five dimensions and gives an unofficial award forecast.
Eight skills for LaTeX research papers, from topic to compiled PDF, with verified BibTeX citations and gated steps
A set of three research skills: paper writing, academic Word and PowerPoint documents, scientific computing and figures
A skill for master and PhD students that runs research through novelty and reality gates, from review to claim checks
A Claude Code skill for research papers: a five-stage pipeline, section rules, a style audit and an independent red-team review.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Math-model-skills (this card) | no | 112 | |
| Math Modeling Skills | no | 437 | offers a set of techniques for modeling tasks; Meta-model-agent fixes seven gated stages and a PDF audit |
| MathModel Skill Standard | no | 458 | sets a solution structure standard; this is a full executable process through to the paper |
README in short
The README describes seven stages, release v1.3 with new model and layout checks and a list of reference documents with a stage routing table.