AI-Research-SKILLs
Comprehensive open-source library of AI research and engineering skills for any AI model. Package the skills and your claude code/codex/gemini agent will be an AI research agent with full horsepower. Maintained by Orchestra Research.
AI-Research-SKILLs is an open-source library of 98 specialized agent skills that enable AI coding assistants to autonomously conduct AI research from idea generation through experimentation to academic paper writing. Maintained by Orchestra Research, it provides deep, production-ready expertise across 23 categories including model architecture, fine-tuning, distributed training, inference, safety, and mechanistic interpretability. Each skill contains comprehensive documentation (300KB+ from official sources), real GitHub issue solutions, code examples, and troubleshooting guides. The library includes an Autoresearch orchestration skill that manages the full research lifecycle and automatically routes to domain-specific skills as needed, enabling truly autonomous research workflows.
Key Features
Use Cases
- 01Autonomous AI research where an agent conducts literature surveys, generates hypotheses, runs experiments, and writes academic papers end-to-end
- 02Fine-tuning large language models using specialized frameworks (Axolotl, LLaMA-Factory, PEFT) with YAML configs and LoRA/QLoRA techniques
- 03Optimizing LLM inference latency and throughput using vLLM, TensorRT-LLM, or SGLang with PagedAttention and quantization
- 04Scaling distributed training to 100+ GPUs using DeepSpeed ZeRO, PyTorch FSDP, or Megatron-Core with 3D/4D parallelism
- 05Conducting mechanistic interpretability experiments with TransformerLens hook points, SAE feature discovery, and causal interventions
- 06Writing publication-ready ML papers for NeurIPS, ICML, ICLR with automated citation verification and LaTeX templates
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
AI-Research-SKILLs — FAQ
What is AI-Research-SKILLs?+
AI-Research-SKILLs is an open-source library of 98 specialized skills that extend AI coding agents (Claude Code, Cursor, etc.) with expert knowledge for conducting autonomous AI research. It covers the full research lifecycle from ideation and literature review through experiments to academic paper writing, with each skill providing comprehensive documentation, code examples, and troubleshooting guides from official sources.
How do I install AI-Research-SKILLs?+
Run 'npx @orchestra-research/ai-research-skills' for an interactive installer that auto-detects your coding agent and installs skills to the appropriate directory (~/.orchestra/skills/ with symlinks). Alternatively, for Claude Code users, add the marketplace with '/plugin marketplace add orchestra-research/AI-research-SKILLs' and install categories like '/plugin install fine-tuning@ai-research-skills'.
Which AI clients and coding agents does it work with?+
The library works with Claude Code, Cursor, Codex, Gemini CLI, Qwen Code, Hermes Agent, OpenCode, OpenClaw, and Qoder. The npm installer automatically detects installed agents and configures the skills appropriately, with fallback to manual copy on Windows systems that don't support symlinks.
Do I need API keys or special prerequisites?+
The skills library itself requires only Node.js/npm for installation. Individual skills may require API keys or accounts depending on the tools they document (e.g., Modal for serverless GPUs, Weights & Biases for experiment tracking, cloud providers for distributed training). Each skill's documentation specifies its own prerequisites.
Is AI-Research-SKILLs free and open source?+
Yes, AI-Research-SKILLs is released under the MIT License and is completely free and open source. However, individual tools and frameworks referenced by the skills may have their own licenses, so check each project's license before use.
What makes the Autoresearch skill different from other skills?+
The Autoresearch skill is a central orchestration layer that manages the entire research lifecycle using a two-loop architecture (inner optimization + outer synthesis). Unlike domain-specific skills, it automatically routes to the appropriate skills for literature surveys, experiments, interpretability analysis, and paper writing, enabling fully autonomous research without human intervention.
How do I install AI-Research-SKILLs?+
Open the source repository on GitHub and follow its README. AI-Research-SKILLs is a skill — MCP Agents Market links you directly to the official repo.
Is AI-Research-SKILLs free?+
AI-Research-SKILLs is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.