autoresearch
Orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid experiment iterations with clear optimization targets. The outer loop synthesizes results, identifies patterns, and steers research direction. Routes to domain-specific skills for execution, supports continuous agent operation via Claude Code /loop and OpenClaw heartbeat, and produces research presentations and papers. Use when starting a research project, running autonomous experiments, or managing a multi-hypothesis research effort.
npx skills add OpenRaiser/NanoResearch --skill 0-autoresearch-skill --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Autoresearch orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid experiment iterations with clear optimization targets. The outer loop synthesizes results, identifies patterns, and steers research direction. Routes to domain-specific skills for execution, supports continuous agent operation via Claude Code /loop and OpenClaw heartbeat, and produces research presentations and papers. Use when starting a research project, running autonomous experiments, or managing a multi-hypothesis research effort.
How it works
You are a research project manager, not a domain expert. You orchestrate; the domain skills execute. This runs fully autonomously. Do not ask the user for permission or confirmation — use your best judgment and keep moving. Show the human your progress frequently through research presentations (HTML/PDF) so they can see what you're doing and redirect if needed. The human is asleep or busy; your job is to make as much research progress as possible on your own.
Getting Started
- Determine user state and proceed accordingly (idea, clear question, plan, or resuming).
- Step 0: set up the agent continuity loop (MANDATORY).
- Initialize workspace structure with folders like research-state.yaml, research-log.md, findings.md, literature/, src/, data/, experiments/, to_human/, paper/.
The Two-Loop Architecture
- BOOTSTRAP: form initial hypotheses after literature search.
- INNER LOOP: pick hypothesis, write protocol, run experiment via domain skill, sanity check, measure, record, learn, next.
- OUTER LOOP: review results, find patterns, update findings.md, form new hypotheses, decide direction, update state, log reflection, and generate progress presentations.
- FINALIZE: write paper and final presentation via ml-paper-writing when concluding.
When to use it
Use when starting a research project, running autonomous experiments, or managing a multi-hypothesis research effort.
What it can touch
Routes to domain-specific skills for execution across data processing, model training, evaluation, inference, interpretability, MLOps, and infrastructure as referenced in the skill’s routing guidance (e.g., 05-data-processing, 01-model-architecture, 11-evaluation, 12-inference-serving, 04-mechanistic-interpretability, 13-mlops, 09-infrastructure).
Caveats
The skill operates autonomously and is designed to run without user permission prompts. It relies on continuous heartbeat mechanisms (Claude Code /loop and OpenClaw) and maintains structured state in files like research-state.yaml, research-log.md, and findings.md. It directs progress through presentations and reports for human oversight.
# Autoresearch Autonomous research orchestration for AI coding agents. You manage the full research lifecycle — from literature survey to published paper — by maintaining structured state, running a two-loop experiment-synthesis cycle, and routing to domain-specific skills for execution. You are a research project manager, not a domain expert. You orchestrate; the domain skills execute. **This runs fully autonomously.** Do not ask the user for permission or confirmation — use your best judgment and keep moving. Show the human your progress frequently through research presentations (HTML/PDF) so they can see what you're doing and redirect if needed. The human is asleep or busy; your job is to make as much research progress as possible on your own. ## Getting Started Users arrive in different states. Determine which and proceed: | User State | What to Do | |---|---| | Vague idea ("I want to explore X") | Brief discussion to clarify, then bootstrap | | Clear research question | Bootstrap directly | | Existing plan or proposal | Review plan, set up workspace, enter loops | | Resuming (research-state.yaml exists) | Read state, continue from where you left off | If things are clear, don'
- Getting Started
- Initialize Workspace
- The Two-Loop Architecture
- Research is Non-Linear
- Bootstrap: Literature and Hypotheses
- The Inner Loop
- Route to Domain Skills
- Track the Experiment Trajectory
- The Outer Loop
- Deciding Direction
- findings.md Is Your Project Memory
- Agent Continuity (MANDATORY — Set Up First)
- Claude Code
- OpenClaw
What does the autoresearch skill do?
Orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid experiment iterations with clear optimization targets. The outer loop synthesizes results, identifies patterns, and steers research direction. Routes to domain-specific skills for execution, supports continuous agent operation via Claude Code /loop and OpenClaw heartbeat, and produces research presentations and papers. Use when starting a research project, running autonomous experiments, or managing a multi-hypothesis research effort.
How do I install it?
Run `npx skills add OpenRaiser/NanoResearch --skill 0-autoresearch-skill --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From OpenRaiser/NanoResearch, a repository with 1,480 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.