A collection of example Jupyter notebooks showing how to build, train, and deploy machine learning models on Amazon SageMaker.
Orchestra-Research/AI-Research-SKILLs
An open-source library of 98 skill packs that teach coding agents like Claude Code, Codex and Gemini CLI to run AI research and engineering tasks.
- Stars
- 13,909
- Forks
- 973
- Open issues
- 23
- Last commit
- 14 days ago
What is it?
AI Research Skills is a collection of markdown-based skill packs for coding agents. Each skill documents a specific AI framework or research stage, such as fine-tuning with Axolotl or LLaMA-Factory, RL post-training with TRL and GRPO, serving with vLLM or SGLang, or quantization with GPTQ and AWQ. The skills are grouped into 23 categories covering model architecture, distributed training, evaluation, RAG, multimodal models, and MLOps.
The library also includes an autoresearch skill that acts as an orchestration layer. It is meant to manage a research workflow from literature survey and ideation through experiments to paper writing, and to route work to the domain skills as needed. Two demo papers in the repository were produced end to end by agents using this skill.
Installation is through an interactive npm command or a Claude Code plugin marketplace. The installer detects supported agents such as Claude Code, Cursor, Gemini CLI and Codex, and links the skills into each one.
Key features
- 01Covers 98 skills across 23 categories, from fine-tuning and inference to evaluation and RAG.
- 02Includes an autoresearch orchestration skill that routes research tasks to domain skills.
- 03Installs with one npm command and detects supported coding agents automatically.
- 04Also available through the Claude Code plugin marketplace, installable by category.
- 05Each skill includes a SKILL.md quick reference plus deeper reference documents.
Use cases
- Giving a coding agent working knowledge of frameworks such as vLLM, TRL or DeepSpeed before a training or serving task.
- Running a multi-stage AI research project where an agent handles literature survey, experiments and paper drafting.
- Looking up a quick reference for a specific ML tool when choosing between alternatives in a category.
Quick start
- 1
Run the interactive installer from the README
npx @orchestra-research/ai-research-skills - 2
List installed skills
npx @orchestra-research/ai-research-skills list
Things to know
- Skills are guidance packs for coding agents, so they need an agent such as Claude Code, Codex or Gemini CLI to be useful.
In collections
Topics
FAQ
- What is AI-Research-SKILLs?
- An open-source library of 98 skill packs that teach coding agents like Claude Code, Codex and Gemini CLI to run AI research and engineering tasks. AI Research Skills is a collection of markdown-based skill packs for coding agents. Each skill documents a specific AI framework or research stage, such as fine-tuning with Axolotl or LLaMA-Factory, RL post-training with TRL and GRPO, serving with vLLM or SGLang, or quantization with GPTQ and AWQ. The skills are grouped into 23 categories covering model architecture, distributed training, evaluation, RAG, multimodal models, and MLOps.
- Is AI-Research-SKILLs free to use? What license does it have?
- AI-Research-SKILLs is open source under the MIT license. Check the license terms before using it in commercial products.
- Is AI-Research-SKILLs still maintained?
- Actively maintained. The last commit to AI-Research-SKILLs was on Oct 11, 2026.
- What can I use AI-Research-SKILLs for?
- Giving a coding agent working knowledge of frameworks such as vLLM, TRL or DeepSpeed before a training or serving task. Running a multi-stage AI research project where an agent handles literature survey, experiments and paper drafting. Looking up a quick reference for a specific ML tool when choosing between alternatives in a category.
- How do I install AI-Research-SKILLs?
- Run the interactive installer from the README: npx @orchestra-research/ai-research-skills List installed skills: npx @orchestra-research/ai-research-skills list
- Where can I get AI-Research-SKILLs?
- The source code and installation instructions are on GitHub: https://github.com/Orchestra-Research/AI-Research-SKILLs
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
Kubernetes-native platform for distributed AI model training and LLM fine-tuning across PyTorch, JAX, XGBoost and other frameworks.
Open-source desktop app and web UI for running, fine-tuning and exporting LLMs and diffusion models locally, plus a code-based training library.
A set of Claude Code skills that walk academic work through research, writing, peer review, revision and finalization, with a human approving each stage.
A cross-tool skill that makes AI coding agents write less code and less prose, cutting output tokens and LLM API costs.
A bilingual collection of 100+ original diagrams explaining LLM, RL, RLHF, DPO, GRPO and related large-model algorithms.