simulation-study

What problem does it solve?

Manually designing and running Monte Carlo simulation studies to evaluate statistical estimator performance is time-consuming, error-prone, and difficult to reproduce. This Skill automates the entire end-to-end workflow, from simulation specification to result validation, ensuring rigorous and reproducible evaluation of finite-sample estimator properties.

Core Features & Use Cases

• Automated Simulation Design: Generates complete sim-spec.md files with data generating process (DGP) definitions, scenario grids, performance metrics, and acceptance criteria tailored to your target estimator.
• Isolated Multi-Pipeline Execution: Coordinates separate code, simulation, and test pipelines to maintain strict isolation between implementation, simulation harness, and validation logic, eliminating bias in simulation results.
• Use Case: For example, if you develop a new robust regression estimator, use this Skill to automatically test its bias, 95% confidence interval coverage, and RMSE across sample sizes from 100 to 5000, with normal and heavy-tailed error distributions.

Quick Start

Use the simulation-study skill to run a Monte Carlo evaluation of the new Huber regression estimator's finite-sample properties.