scikit-learn

multiscale-materials-modelling

A Claude Code skill marketplace for multiscale materials modelling workflows. Each of the 22 skills installs individually — pick only the ones you need. Built for the SuperScientist agent framework, it covers molecular dynamics, quantum chemistry, machine learning potentials, and materials informatics — from building a simulation box to analyzing production trajectories.

Features

• Molecular dynamics — LAMMPS, GROMACS, and packmol for setting up and running atomistic and coarse-grained simulations
• Quantum chemistry — PySCF for HF/DFT/MP2 calculations and CP2K for periodic DFT and ab initio MD
• Machine learning potentials — MACE for fast, accurate energy and force predictions
• Materials informatics — Pymatgen and the Materials Project API for crystal structures, phase diagrams, and properties
• Cheminformatics — RDKit for SMILES parsing, descriptors, and substructure searches
• Analysis and visualization — ASE, freud, mdanalysis, matplotlib, and seaborn for trajectories and plots
• Optimization — Bayesian optimization for composition and property tuning via Ax/BoTorch
• Skill creator tooling — Built-in guidance for adding new multiscale materials modelling skills

Skills

Group	Skill	Purpose
Structure & I/O	ase	Atomistic structure manipulation, format conversion, geometry optimization
Structure & I/O	lammpsio	Read/write LAMMPS data and dump files
Structure & I/O	packmol	Build initial configurations for MD simulations
Structure & I/O	polymer-build	Generate polymer systems and LAMMPS data files from SMILES
Molecular Dynamics	lammps	Write, review, and debug LAMMPS input scripts
Molecular Dynamics	gromacs	Biomolecular MD setup, execution, and analysis with GROMACS
Quantum Chemistry	pyscf	Molecular HF, DFT, and MP2 calculations in Python
Quantum Chemistry	cp2k	Periodic DFT, AIMD, and geometry optimization with CP2K
Machine Learning	mace	Run pretrained MACE-MP models for energies and forces
Materials Science	pymatgen	Crystal structures, format conversion, phase diagrams
Materials Science	materials-project	Query the Materials Project database
Cheminformatics	rdkit	SMILES parsing, descriptors, fingerprints, substructure search
Analysis	freud	RDF, MSD, order parameters, PMFT, clustering for trajectories
Analysis	matplotlib	General-purpose plotting and figure customization
Analysis	scientific-plotting	Publication-quality figures with journal styling (APS, Nature, Science)
Analysis	ovito	Render molecular snapshots and animations from trajectories
Analysis	seaborn	Statistical visualizations from DataFrames
Analysis	shap	Model-agnostic feature importance and explanations
Analysis	scikit-learn	ML models, clustering, and dimensionality reduction
Analysis	umap-learn	Nonlinear dimensionality reduction and embedding visualization
Optimization	bayes-opt	Bayesian optimization for expensive material properties
Meta	skill-creator	Create, edit, and evaluate new skills

When to Use Which Tool

Task	Recommended Skill
Build a polymer melt and equilibrate with LAMMPS	polymer-build → packmol → lammps
Run a protein-ligand simulation	gromacs
Compute a molecule's DFT energy or optimize its geometry	pyscf (gas phase) or cp2k (periodic)
Screen catalysts with a neural-network potential	mace
Get crystal structures and compute a phase diagram	materials-project → pymatgen
Analyze RDF or diffusion from a trajectory	freud or mdanalysis
Optimize a multi-component formulation	bayes-opt
Convert between CIF, POSCAR, XYZ, LAMMPS data	ase or pymatgen

Installation

Add the marketplace once:

/plugin marketplace add Chenghao-Wu/cc-skills-ZhenghaoWu-Group

Then install only the skills you need — each skill is its own plugin:

/plugin install lammps@multiscale-materials-modelling-marketplace
/plugin install rdkit@multiscale-materials-modelling-marketplace
/plugin install freud@multiscale-materials-modelling-marketplace

Or run /plugin to browse and install skills interactively. After installing, restart Claude Code (or run /reload-plugins) so the new skills load.

Python dependencies

A skill assumes its underlying library is available in your Python environment (for example freud needs freud-analysis, rdkit needs rdkit). Install what each skill needs, or install everything at once for development and testing:

git clone https://github.com/Chenghao-Wu/cc-skills-ZhenghaoWu-Group.git
cd cc-skills-ZhenghaoWu-Group
pip install -e ".[dev]"