Convert problem attachments into reproducible end-to-end analysis pipelines.
npx skills add https://github.com/twj0/2026mcm --skill mcm-c-coder
Please help me install this Agent Skill. Skill: mcm-c-coder Source: https://github.com/twj0/2026mcm/tree/main/.windsurf/skills/mcm-c-coder Command: npx skills add https://github.com/twj0/2026mcm --skill mcm-c-coder
Coding specialist for COMAP MCM/ICM Problem C (C题数据处理与建模落地). Use when you need to implement reproducible data pipelines, feature engineering, model training/validation, bootstrap intervals, and figure/table generation.
Run the end-to-end pipeline on the provided problem data to reproduce the analysis and generate publication-ready figures and tables.
High-intent search queries and answers about installing and using this skill.
A reproducible data pipeline for MCM Problem C converts problem attachments into an end-to-end workflow: reading, cleaning, feature engineering, model training, evaluation, and visualization. Scripts and configurations ensure identical results across iterations.
Yes, a uv-based environment manager is required to run this Python data analysis pipeline. It manages the Python stack including pandas, numpy, scipy, and scikit-learn, ensuring strict no-leak validation and consistent reproducibility across environments.
The best way to generate publication-ready figures and tables from MCM C data is through an end-to-end pipeline using matplotlib and seaborn. It produces robust evaluation outputs, bootstrap intervals, and visuals suitable for publication.
Strict no-leak validation during model training and evaluation prevents data leakage across training and testing sets. This mechanism ensures reproducible bootstrap intervals and robust evaluation metrics for consistent publication-quality results.
Yes, statsmodels is optionally supported for feature engineering within this MCM data workflow. The pipeline integrates it alongside pandas, numpy, scipy, and scikit-learn to process problem attachments and train models.
Different results across iterations indicate a lack of reproducibility in your MCM C data pipeline. Using a uv-based environment with strict no-leak validation and fixed configurations ensures identical results and consistent outputs.
Yeachan-HeoLoad, clean, and summarize tabular datasets with descriptive statistics.
llv22Generate publication-quality figures and LaTeX-ready tables from experimental data.
tassiovaleGenerate publication-ready scientific figures with matplotlib, seaborn, and plotly.
pluginagentmarketplaceAutomate end-to-end machine-learning workflows with scikit-learn and XGBoost.
