fundamental-physics

Fundamental Physics Plugins

Claude Code plugins for fundamental physics analysis, with an initial focus on astrophysics and cosmology.

Overview

This repository provides skills, tools, and commands for scientific workflows in fundamental physics research. Plugins integrate with Claude Code to assist with:

• Astronomical data processing and analysis
• Cosmological computations
• Literature search and management (arXiv, INSPIRE-HEP, NASA ADS)
• Scientific Python workflows

Available Plugins

astro-ph.CO

Astrophysics analysis workflows, scientific computing, and astronomical data processing.

Skills:

• astrophysics - Observational astronomy, cosmological analysis, FITS/HDF5 data handling, spectroscopy, photometry

core

Core tools for fundamental physics research: literature access, data repositories, and paper management.

Skills:

• arxiv - Search arXiv and download paper source (LaTeX/BibTeX)
• inspire - INSPIRE-HEP search, citations, BibTeX for high-energy physics
• ads - NASA ADS search, citations, BibTeX for astronomy/astrophysics (requires API token)
• zenodo - Search and download datasets, software, and research artifacts with DOIs
• hepdata - Download experimental data tables from HEP publications (cross-sections, limits, etc.)
• science-skill-creator - Guide for creating effective skills for fundamental physics research

Dependencies: requests, pyyaml (pip install requests pyyaml)

gr-qc

General relativity and quantum cosmology: gravitational wave data and related tools.

Skills:

• gwosc - Search gravitational wave events, get parameters, download strain data from LIGO/Virgo/KAGRA

Dependencies: requests (pip install requests)

Installation

Within Claude Code install directly with:

/plugin marketplace add git@github.com:fundamental-physics/marketplace.git

Creating Your Own Physics Skills

Want to create a custom skill for your research workflow? Use the science-skill-creator:

1. Ask Claude: "Help me create a new skill for [your physics workflow]"
2. The science-skill-creator will guide you through:
   • Understanding your use case with concrete examples
   • Planning reusable components (scripts, references, assets)
   • Implementing the skill following physics research patterns
   • Testing and iterating on the skill

Example usage:

• "Help me create a skill for running JAX-based nested sampling on my cosmology models"
• "I need a skill for processing LIGO gravitational wave data"
• "Create a skill for lattice QCD simulations with gauge field configurations"

The skill creator follows a physics workflow model (Research → Science → Inference → Visualization) and helps you build skills that integrate seamlessly with your research pipeline.

Repository Structure

plugins/
└── <plugin-name>/
    ├── .claude-plugin/
    │   └── plugin.json       # Plugin manifest
    ├── skills/
    │   └── <skill-name>/
    │       ├── SKILL.md      # Skill documentation
    │       └── scripts/      # Optional: bundled scripts
    ├── commands/             # Optional: slash commands
    └── agents/               # Optional: subagents

Contributing

Adding a New Skill to an Existing Plugin

If your skill fits within an existing plugin's domain, add it there:

1. Create a new directory: plugins/<plugin>/skills/<skill-name>/
2. Add a SKILL.md file with YAML frontmatter and documentation
3. Optionally add supporting scripts in scripts/

Creating a New Plugin

Plugins are organized by arXiv category (e.g., astro-ph.CO, hep-th, gr-qc). To add a new plugin:

1. Create a directory under plugins/ named after the arXiv category
2. Add skills, commands, or agents as needed
3. Register the plugin in .claude-plugin/marketplace.json by adding an entry to the plugins array:

   {
     "name": "your-category",
     "source": "./plugins/your-category"
   }
   

Skill Review Requirements

All new skills must be reviewed against the official Claude Code skills documentation:

https://code.claude.com/docs/en/skills

Reviewers should verify:

• Correct YAML frontmatter structure (name, description, allowed-tools)
• Clear trigger conditions in the description
• Appropriate tool permissions
• Well-documented usage examples
• Following progressive disclosure (overview first, details later)

License

MIT