experiment

Create, list, and remove automated code improvement experiments with JSON and Git worktrees.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/tokyo-megacorp/autoimprove --skill experiment-tokyo-megacorp
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: experiment
Source: https://github.com/tokyo-megacorp/autoimprove/tree/main/skills/experiment
Command: npx skills add https://github.com/tokyo-megacorp/autoimprove --skill experiment-tokyo-megacorp

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill simplifies the process of creating, listing, and removing autoimprove experiments, streamlining experiment management for iterative code improvement.

Core Features & Use Cases

  • Create experiments: Guided wizard to define themes, hypotheses, target files, constraints, and expected outcomes.
  • List experiments: View detailed experiment records with filters for status, theme, and date.
  • Remove experiments: Interactive selection and confirmation to delete experiment data and associated resources.

Quick Start

Use the experiment skill to create a new experiment and manage existing ones efficiently.

Frequently Asked Questions about experiment

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I manage automated code improvement experiments in Git?

To manage automated code improvement experiments, you can use a Skill that handles directory operations, JSON configuration, and Git worktree operations to create, track, and delete experiments efficiently.

What is the best way to track iterative testing and hypothesis validation for code improvements?

Tracking iterative testing and hypothesis validation is done by creating experiments through a guided wizard that defines themes, target files, constraints, and expected outcomes for comprehensive management.

How do I create and configure a new automated experiment for my development workflow?

To create and configure a new automated experiment, you use a guided wizard to define the experiment's theme, hypothesis, target files, constraints, and expected outcomes, saving the configuration via JSON.

Can I filter and list existing automated experiments by status or date?

Yes, you can list existing automated experiments and view detailed records using filters for status, theme, and date to easily track your development workflow.

Does the experiment management process require Git worktree operations?

Yes, comprehensive experiment management implements Git worktree operations alongside TSV parsing and directory handling to ensure full control over automated code improvement experiments.

How do I safely remove automated experiments and their associated resources?

To safely remove automated experiments, you use an interactive selection and confirmation process that deletes the experiment data along with all associated resources.