What is the mcaf-ml-ai-delivery skill?

It is a specialized Claude Code skill designed to provide domain-specific guidance for the delivery of machine learning and AI systems, covering everything from data exploration to model operations.

Can this skill generate project documentation?

Yes, it produces concrete artifacts including clearer delivery guidance, links between experimentation and verification, and documentation that matches the actual system architecture.

How does this skill handle Responsible AI?

It treats Responsible AI as a first-class requirement, ensuring that ethics, data quality, and verification steps are integrated into the workflow rather than added as an afterthought.

When should I use this skill instead of generic developer tools?

Use this skill specifically when your repository contains model training, inference, or data science workflows that require explicit process and responsible-AI guidance.

What is the 'Ralph Loop' in this context?

The Ralph Loop is a mandatory iterative planning and execution cycle used by the skill to define targets, execute changes, and validate results through concrete evidence and documentation.

ML/AI Delivery Lifecycle Management

Name: ML/AI Delivery Lifecycle Management
Author: managedcode

bymanagedcode

•

データサイエンスとML

Guides the end-to-end delivery of ML/AI projects through structured experimentation, data exploration, and responsible AI practices.

The mcaf-ml-ai-delivery skill provides a comprehensive framework for managing the complexities of machine learning and artificial intelligence project lifecycles within Claude Code. It streamlines the transition from initial data exploration and feasibility studies to model training, inference, and MLOps. By enforcing a rigorous iterative planning process known as the Ralph Loop, the skill ensures that product, data, and model assumptions are clearly separated, while integrating responsible AI and data quality checks directly into the development workflow. This results in highly traceable experiments, validated model artifacts, and professional documentation.

主な機能

01Traceable experimentation tracking and evaluation documentation

02Separation of product, data, and model logic to reduce technical debt

031 GitHub stars

04Integrated Responsible AI and data quality verification frameworks

05Iterative 'Ralph Loop' workflow for planning and validation

06Structured guidance for data exploration and feasibility studies

ユースケース

01Implementing and documenting Responsible AI checks for production models

02Establishing a standardized delivery workflow for new machine learning features

03Organizing complex data science repositories with clear separation of concerns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add managedcode/mcpgateway mcaf-ml-ai-delivery

For use in Claude.ai and ChatGPT

主な機能

01Traceable experimentation tracking and evaluation documentation

02Separation of product, data, and model logic to reduce technical debt

031 GitHub stars

04Integrated Responsible AI and data quality verification frameworks

05Iterative 'Ralph Loop' workflow for planning and validation

06Structured guidance for data exploration and feasibility studies

ユースケース

01Implementing and documenting Responsible AI checks for production models

02Establishing a standardized delivery workflow for new machine learning features

03Organizing complex data science repositories with clear separation of concerns