# Python Model Code Generator

Generate and run minimal reproducible Python modeling code for the human-approved main...

- Repository: https://github.com/zhnnky329/MathModeling-skills (1073 stars)
- Category: Engineering
- License: MIT
- Last updated: 2026-09-24
- Runs scripts: no, instructions only

## Install

Claude Code: `npx skills add zhnnky329/MathModeling-skills --skill python-model-code-generator -g -a claude-code -y`
Codex: `npx skills add zhnnky329/MathModeling-skills --skill python-model-code-generator -g -a codex -y`
Cursor: `npx skills add zhnnky329/MathModeling-skills --skill python-model-code-generator -g -a cursor -y`
Gemini CLI: `npx skills add zhnnky329/MathModeling-skills --skill python-model-code-generator -g -a gemini-cli -y`
Claude Code plugin marketplace: `/plugin marketplace add zhnnky329/MathModeling-skills`

## Description

Generate and run minimal reproducible Python modeling code for the human-approved main method and usable baseline, saving compact experiment artifacts and a canonical run summary.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Preconditions

- G2.5 human method choice is recorded in `methods/Qx/qx_decisions.jsonl`.
- `code/Qx/qx_code_plan.md` exists.
- Required cleaned data and profile exist.
- The plan targets Python.

Legacy method pools and `code/model-code-analyzer.md` may be read during migration, but they do not override the human choice.

# Workflow

1. Read the code plan, decision ledger, method card, probe conditions, and data profile.
2. Confirm scope:
   - one approved main method;
   - one usable baseline;
   - fallback only when an activation decision or evidenced trigger exists.
3. Generate clear runnable `.py` files under `code/Qx/`.
4. Use project-root-safe paths, fixed seeds, explicit inputs, and minimal justified dependencies.
5. Save:
   - tables to `results/Qx/experiments/roundN/tables/`;
   - metrics to `.../metrics/`;
   - useful diagnostic/comparison figures to `.../figures/`;
   - canonical `run_summary.json`.
6. Evaluate and record output-degeneracy and fallback-trigger metrics required by the plan.
7. Persist full logs only on failure or when a warning needs reproduction.
8. Run the code. Do not claim success from code generation alone.
9. Hand off to `code-reviewer`.

# Script Layout

Prefer the smallest clear layout:

```text
code/Qx/
├── qx_code_plan.md
├── qx_baseline.py
├── qx_main.py
└── run_all.py        # only when coordination is useful
```

Do not create one script per unapproved candidate. Do not create a README that duplicates the code plan.

# Run Summary

Follow the schema in `model-code-analyzer`. Include:

- approved decision ID;
- method IDs and roles;
- inputs and outputs;
- seed and environment;
- execution status and timing;
- compact metric summaries;
- output-degeneracy evidence;
- warnings/errors;
- fallback-trigger state.

# Rules

The rest: https://github.com/zhnnky329/MathModeling-skills/blob/HEAD/.claude/skills/python-model-code-generator/SKILL.md