computational-experiments

What problem does it solve?

This Skill automates the lifecycle of computational research experiments, from scaffolding to publication output, saving researchers time and reducing errors.

Core Features & Use Cases

• Scaffolding: Automates the creation of project structures, code skeletons, and test stubs.
• Experiment Design: Manages sweep campaigns, adaptive exploration, and autonomous self-correcting sweeps.
• Publication Output: Generates convergence plots, comparison charts, ablation plots, sensitivity analysis, and tables.
• Use Case: Imagine you're running a series of experiments on a new algorithm. Use this Skill to automatically run the experiments, generate publication-ready figures, and write a reproducibility report.

Quick Start

Use the computational-experiments skill to run a full experiment with a given configuration and budget: compute-experiments run --config my-experiment.yaml --budget 30