A set of Claude Agent SDK skills for AI-powered epidemiological analysis using the Epiverse-TRACE R ecosystem.
Epiagent Skills gives Claude the ability to perform infectious disease epidemiology analyses — from outbreak simulation to writing a Quarto report — by providing structured skill instructions grounded in the Epiverse-TRACE package ecosystem.
This repository contains only skills for the Claude Agent SDK. There is no standalone application code. The skills live in .claude/skills/ and are loaded by Claude Code when working in this directory.
epiagent_skills/
└── .claude/skills/
├── epiverse_overview/ # Orchestrator for routing tasks
├── data_intake/ # Data cleaning and standardization
├── parameters/ # Epidemiological parameter management
├── simulation/ # Outbreak simulation
├── analysis/ # Statistical analysis (CFR, Rt, etc.)
├── visualisation/ # Publication-ready plots
└── reporting/ # Publication-ready Quarto reports
Orchestrator skill that routes tasks to the appropriate specialized skill. Start here for task planning.
Clean and standardize epidemiological data using readepi, cleanepi, linelist, and numberize.
Retrieve and manage epidemiological parameters (incubation periods, serial intervals) using epiparameter and epiparameterDB.
Generate synthetic outbreak data — line lists, contact data, transmission chains — using simulist and epichains.
Estimate CFR, Rt, final size, superspreading, and vaccine effectiveness using cfr, finalsize, epidemics, superspreading, and vaccineff.
Create publication-ready plots with the Epiverse-TRACE branded theme using ggplot2 and tracetheme.
Generate publication-ready Quarto reports (HTML, PDF, Word) with standard academic structure.
Each skill maps to a stage in the epidemiology pipeline:
Data Intake → Parameters → Simulation → Analysis → Visualisation → Report
Skills can be used independently or chained together for end-to-end workflows.
To use these skills, you need R (≥ 4.3.0) and the Epiverse-TRACE packages installed:
options(repos = c(
epiverse = "https://epiverse-trace.r-universe.dev",
CRAN = "https://cloud.r-project.org"
))
install.packages("pacman")
pacman::p_load(
simulist, cleanepi, linelist, cfr, epiparameter,
finalsize, epidemics, superspreading, vaccineff, epichains,
readepi, epiparameterDB, tracetheme, numberize,
tidyverse, here, qs, ggplot2
)Then clone this repository into your working directory so Claude Code can load the skills.
MIT License — see LICENSE for details.
Skills for Claude Agent SDK