ParticlePhysics MCP Server
Provides seamless access to particle physics data from the Particle Data Group (PDG) for AI assistants and applications.
ParticlePhysics MCP Server
A Model Context Protocol server that lets Claude Desktop, IDEs, and other MCP clients look up particle properties and decay modes.
Supports natural-language queries (muon plus, pion zero, antiproton, anti up quark),
case-insensitive lookup, MC IDs (-13), and returns both human-readable text and a
structured JSON payload.
Table of Contents
Features
search_particle— mass (MeV + GeV), charge, spin, color, parity / C / I / G, lifetime or width, MCID, review IDlist_decays— exclusive / inclusive branching fractions with the source method named- Natural-language input —
muon plus,positive tau,pion zero,kaon minus - Anti-particle support —
antimuon,anti up quark,ubar,u bar,u_bar,u~,antineutron; resolved via MCID negation, no name guessing - MC ID lookup — query directly with
11,-2212, etc. - Self-conjugate aware —
anti photonresolves togamma,anti pi0topi0 - Structured output — every response includes a fenced
```jsonblock alongside the human-readable text
Install
git clone https://github.com/uzerone/particlephysics-mcp-server.git
cd particlephysics-mcp-server
pip install -e .
Configure your MCP client
Add one of the following to your client's MCP config (e.g. claude_desktop_config.json).
Using uvx from the cloned repo (no global install needed):
{
"mcpServers": {
"particlephysics": {
"command": "uvx",
"args": ["--from", "/absolute/path/to/particlephysics-mcp-server",
"python", "-m", "particlephysics_mcp_server"]
}
}
}
Using a local virtualenv (after pip install -e .):
{
"mcpServers": {
"particlephysics": {
"command": "/absolute/path/to/.venv/bin/python",
"args": ["-m", "particlephysics_mcp_server"]
}
}
}
Tools
search_particle
| Input | Example |
|---|---|
| Canonical name | mu+, pi0, K-, Sigma+, gamma, Lambda, H |
| English alias | muon, pion, higgs, electron, top quark |
| Anti-particle | antimuon, anti up quark, ubar, u bar, u_bar, u~, antiproton |
| Natural-language charge | muon plus, positive tau, pion zero, kaon minus |
| MC ID | 11 (electron), -13 (mu+), 2212 (proton) |
Sample call: search_particle({"query": "muon plus"})
Found 1 particle(s) matching 'muon plus':
1. mu
Name: mu+
PDG ID: -13
PDG Review ID: S004/2025
Mass: 105.6583755 MeV (0.1056583755 GeV)
Spin (J): 1/2
Charge: 1
Color: singlet
Quantum numbers: J=1/2
Lifetime: 2.196981148893498e-06
```json
{
"query": "muon plus",
"count": 1,
"particles": [{
"name": "mu+",
"mcid": -13,
"pdg_review_id": "S004/2025",
"mass": {"mev": 105.6583755, "gev": 0.1056583755},
"charge": {"value": 1.0, "fraction": "1"},
"spin": "1/2",
"color": {"multiplicity": 1, "label": "singlet"},
"quantum_numbers": {"J": "1/2"},
"lifetime": {"seconds": 2.197e-06, "stable": false, "text": "..."}
}]
}
```
list_decays
Same identifier formats as search_particle. Tries exclusive_branching_fractions →
branching_fractions → inclusive_branching_fractions and reports which source it used.
Returns an empty decay list with stable=true for stable particles.
Sample call: list_decays({"particle_id": "tau"})
Decay modes for particle 'tau':
1. tau- --> mu- nubar_mu nu_tau (BR: 17.39 ± 0.04 %)
2. tau- --> e- nubar_e nu_tau (BR: 17.82 ± 0.04 %)
…
```json
{
"particle": {"name": "tau-", "mcid": 15, ...},
"source": "exclusive_branching_fractions",
"count": 137,
"decays": [{
"description": "tau- --> mu- nubar_mu nu_tau",
"branching_ratio_text": "17.39 ± 0.04",
"value_text": "17.39E-2",
"is_limit": false
}, ...]
}
```
Claude Skill
A Claude Code skill spec lives at .github/skills/particlephysics-skill/SKILL.md. It runs the inspector, validates both tools, and exercises the natural-language / anti-particle / MC ID query surface.
Trigger phrase: particle physics mcp or pp.
Changelog
See CHANGELOG.md for release notes.
Maintainer
License
MIT — see LICENSE.txt.
Servidores relacionados
MongoDB That Works
A MongoDB MCP server with schema discovery and field validation. Requires a MONGODB_URI environment variable.
Bamwor World Data
Access data on 261 countries and 13.4 million cities — population, GDP, geography, rankings, and comparisons. Built for Claude, Cursor, and AI agents.
Data Pilot (Snowflake)
A comprehensive Model Context Protocol (MCP) server for interacting with Snowflake using natural language and AI.
CData Square Server
A read-only MCP server for querying live data from Square using the CData JDBC Driver.
NY Benchmark
Query 2M+ municipal finance data points across New York State — 62 cities, 57 counties, 689 school districts. 30 years of audited actuals with domain-aware caveats applied automatically.
Theta Health MCP Server
Connect your health data to AI assistants like Cursor, Claude, and Windsurf.
Alpha Vantage MCP Server
Enables AI agents and applications to access real-time and 20+ years historical financial market data through natural language queries.
Fantasy Premier League
Access Fantasy Premier League (FPL) data and tools, including player information, team details, and gameweek data.
Sanity MCP Server
Connects Sanity projects with AI tools, allowing AI models to understand content structure and perform operations using natural language.
UniProt MCP Server
Fetch protein information from the UniProt database.