How does the discover-ml skill activate?

It activates automatically when Claude Code detects keywords related to machine learning development, such as 'PyTorch', 'inference', 'training', or 'model selection'.

What specialized ML sub-skills are included?

The gateway provides access to 25 skills, including RAG evaluation, DSPy modules, LoRA/PEFT techniques, Stable Diffusion deployment, and LLM-as-judge frameworks.

How do I see the full list of available ML workflows?

Once the gateway is active, you can run 'cat ~/.claude/skills/ml/INDEX.md' to view the complete list of descriptions, triggers, and workflow combinations.

How does progressive loading benefit my development?

Progressive loading prevents context bloat by only loading the 'Gateway' initially. You can then load the category index or specific skill files only when needed.

Can I use this for both LLM and Image Generation tasks?

Yes, it includes skills for both Large Language Model (LLM) workflows like RAG and DSPy, as well as diffusion model basics and fine-tuning.

Machine Learning Discovery Gateway

Name: Machine Learning Discovery Gateway
Author: rand

byrand

•

数据科学与机器学习

Automatically identifies and activates specialized machine learning and AI development skills within Claude Code.

The discover-ml skill acts as a comprehensive gateway for AI development, automatically activating whenever machine learning, model training, or inference tasks are detected. It provides organized access to a structured library of 25 specialized sub-skills covering critical domains such as RAG architectures, DSPy optimization, LLM evaluation frameworks, and fine-tuning techniques for diffusion and language models. By utilizing a progressive loading system, this skill ensures Claude remains context-efficient while maintaining instant access to high-quality implementation patterns and domain-specific best practices.

主要功能

01Comprehensive indexing of fine-tuning, deployment, and benchmarking patterns

02Automatic activation for ML, AI, and model development tasks

03Support for PyTorch, TensorFlow, and Unsloth optimization workflows

04Access to 25 specialized skills including RAG, DSPy, and LLM evaluation

05Progressive loading system to maintain context window efficiency

0657 GitHub stars

使用场景

01Optimizing LLM application logic using DSPy signatures and modules

02Building and evaluating advanced RAG pipelines with hybrid search and reranking

03Implementing model fine-tuning and evaluation using PEFT and LoRA techniques

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rand/cc-polymath discover-ml

For use in Claude.ai and ChatGPT

Download Skill