A curated collection of reusable Agent Skills for AI coding assistants such as Claude Code, OpenCode, Hermes Agent, Cursor, and other LLM-powered development tools.
This repository is forked from David Ondrej's Agent Skills and extended with additional custom skills, improvements, and workflows that I personally use for software engineering, machine learning, AI development, and research.
The goal is to build a practical library of reusable instructions that help AI agents complete tasks more consistently and efficiently.
Compared to the original repository, this fork includes:
- ✅ Additional custom Agent Skills
- ✅ Improved workflows
- ✅ ML & Deep Learning focused skills
- ✅ AI engineering utilities
- ✅ Research and documentation workflows
- ✅ Continuous additions as I develop new workflows
skills/
├── agent-orchestration/
├── ops-and-setup/
├── research-and-web/
├── skill-authoring/
├── thinking-and-docs/
└── ...
Every skill is self-contained and includes a SKILL.md file describing:
- Purpose
- When to use it
- Required inputs
- Workflow
- Expected outputs
- Best practices
Coordinate multiple AI agents, delegate tasks, automate workflows, and manage agent execution.
Gather information from the web, summarize documentation, extract knowledge, and perform technical research.
Generate documentation, explain concepts, brainstorm ideas, write tutorials, and organize technical knowledge.
Development environment setup, deployment workflows, server configuration, security, and developer tooling.
Create, organize, and improve reusable Agent Skills.
Custom skills focused on:
- Model training
- Dataset preparation
- Deep Learning
- Computer Vision
- NLP
- Experiment tracking
- MLOps workflows
- Kaggle competitions
This repository serves as a growing library of reusable workflows that enable AI assistants to:
- Write better code
- Debug faster
- Automate repetitive engineering tasks
- Conduct technical research
- Produce higher-quality documentation
- Improve software development productivity
- Assist with Machine Learning and AI projects
Planned additions include:
- More ML engineering skills
- Deep Learning workflows
- Computer Vision pipelines
- NLP workflows
- MLOps automation
- Data Engineering utilities
- Software architecture skills
- Prompt engineering techniques
This project is based on the excellent Agent Skills repository created by David Ondrej.
Original repository: https://github.com/davidondrej/skills
Many thanks to David for creating and open-sourcing the original project.
This repository follows the license of the original project unless otherwise specified.