ml-iterate

Automates hypothesis-driven ML experiment iteration workflows with structured documentation and tracking.

1|Updated Jan 3, 2022

Search Tags:#workflow #documentation #hypothesis #ml #experiments #experiment-design #readmes

Ready to run

npx skills add https://github.com/daikichiba9511/dotfiles --skill ml-iterate-daikichiba9511

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ml-iterate
Source: https://github.com/daikichiba9511/dotfiles/tree/main/private_dot_codex/skills/ml-iterate
Command: npx skills add https://github.com/daikichiba9511/dotfiles --skill ml-iterate-daikichiba9511

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

ML teams often run hypothesis-driven experiments in iterative cycles, but manually coordinating ideas, experiments, and documentation leads to slow progress and missed learnings.

Core Features & Use Cases

Phase-driven hypothesis to experiment cycle with guided README updates.
Automatic tracking of related experiments and their outcomes to build a knowledge base.
Use Case: A data science team wants to formalize experiments for model selection and compare results across iterations.

Quick Start

Run /ml-iterate <exp_dir> to start or continue an experiment, e.g., /ml-iterate exp003.

Frequently Asked Questions about ml-iterate

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

FAQPage Schema

How do I track machine learning hypotheses across multiple experiment iterations?▼

Tracking machine learning hypotheses requires linking phase definitions to experiment designs. This workflow organizes hypotheses and records results, maintaining a traceable knowledge base for model selection across iterations.

What is a hypothesis-driven ML experiment workflow?▼

A hypothesis-driven ML experiment workflow is a structured cycle of defining phases, designing tests, and documenting results. It replaces manual coordination by automatically tracking related experiments and outcomes to build a knowledge base.

How do I document machine learning experiments to ensure traceable results?▼

To document machine learning experiments for traceable results, use a phase-driven cycle that guides README updates and automatic tracking. This structured reporting connects experiment designs to outcomes across iterations.

Can I use this to formalize model selection experiments for a data science team?▼

Yes, data science teams can use this workflow to formalize experiments for model selection. It supports explicit phase definitions and structured reporting, allowing you to compare results across iterations.

How do I start an ML experiment iteration using a directory structure?▼

You start an ML experiment iteration by pointing the workflow to your target directory. This initiates or continues the experiment, applying guided README updates and tracking the cycle from hypothesis to result.