What information does the generated PRD include?

The PRD includes the project name, problem statement, success criteria (like AUC or training time), hardware constraints, and the specific scope of the experiment.

What is the primary purpose of ML-Ralph?

ML-Ralph acts as an architect to help you define the requirements for a machine learning project and then hand those requirements off to an autonomous agent for execution.

How do I start the PRD creation process?

You can trigger the skill by using keywords such as 'ml-ralph', 'create prd', or 'ml project' within your Claude Code environment.

Where are the project requirements saved?

The skill automatically creates and writes all project details to a file located at .ml-ralph/prd.json in your local directory.

Can ML-Ralph help with Kaggle competitions?

Yes, it is specifically designed to handle the fast-paced requirements of Kaggle, including defining validation strategies like cross-validation or time splits.

ML-Ralph Project Architect

Name: ML-Ralph Project Architect
Author: pentoai

bypentoai

•

数据科学与机器学习

Generates comprehensive Product Requirements Documents (PRDs) for machine learning experiments and Kaggle competitions through guided conversation.

ML-Ralph is a specialized skill designed to streamline the initialization of machine learning workflows. By facilitating a structured dialogue regarding objectives, metrics, data constraints, and validation strategies, it transforms high-level project ideas into actionable .ml-ralph/prd.json configuration files. This skill is ideal for data scientists and ML engineers who need to quickly define experimental scopes, identify potential data leakage risks, and establish clear success criteria before handing off tasks to autonomous agents for model training and optimization.

主要功能

01Risk assessment for data leakage and compute limitations

02Automated generation of standardized .ml-ralph/prd.json files

03Guided conversational discovery for ML project requirements

04Identification of success metrics, constraints, and validation strategies

0530 GitHub stars

06Seamless handoff to autonomous ML experimentation agents

使用场景

01Scoping experimental data science features within existing software repositories

02Initializing a Kaggle competition entry with defined success metrics and constraints

03Creating structured PRDs for production-level machine learning model development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pentoai/ml-ralph ml-ralph

For use in Claude.ai and ChatGPT

Download Skill