Autoresearch
Autonomous iterative experimentation loop for any programming task. Guides the user through defining goals, measurable metrics, and scope constraints, then runs an autonomous loop of code changes, testing, measuring, and keeping/discarding results. Inspired by Karpathy's autoresearch. USE FOR: autonomous improvement, iterative optimization, experiment loop, auto research, performance tuning, automated experimentation, hill climbing, try things automatically, optimize code, run experiments, autonomous coding loop. DO NOT USE FOR: one-shot tasks, simple bug fixes, code review, or tasks without a
Install
npx skills add a-tokyo/aiworkspace --skill autoresearchgit clone --depth 1 https://github.com/a-tokyo/aiworkspace.git /tmp/aiworkspace
mkdir -p ~/.claude/skills
cp -r /tmp/aiworkspace/.agents/skills/autoresearch ~/.claude/skills/autoresearchPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# Autoresearch: Autonomous Iterative Experimentation An autonomous experimentation loop for any programming task. You define the goal and how to measure it; the agent iterates autonomously -- modifying code, running experiments, measuring results, and keeping or discarding changes -- until interrupted. This skill is inspired by [Karpathy's autoresearch](https://github.com/karpathy/autoresearch), generalized from ML training to **any programming task with a measurable outcome**. --- ## Agent Behavior Rules 1. **DO** guide the user through the Setup phase interactively before starting the loop. 2. **DO** establish a baseline measurement before making any changes. 3. **DO** commit every experiment attempt before running it (so it can be reverted cleanly). 4. **DO** keep a results log (TSV) tracking every experiment. 5. **DO** revert changes that do not improve the metric (git reset to last known good). 6. **DO** run autonomously once the loop starts -- never pause to ask "should I continue?". 7. **DO NOT** modify files the user marked as out-of-scope. 8. **DO NOT** skip the measurement step -- every experiment must be measured. 9. **DO NOT** keep changes that regress the metric unless the user explicitly allowed trade-offs. 10. **DO NOT** install new dependencies or make environment changes unless the user approved it. --- ## Phase 1: Setup (Interactive) Before any experimentation begins, work with the user to establish these parameters.
Details
- Skill name
autoresearch- Repository
- a-tokyo/aiworkspace
- Path
.agents/skills/autoresearch/SKILL.md- License
- Apache-2.0
- Repo updated
- 19 August 2026
FAQ
What does the Autoresearch skill do?
Autonomous iterative experimentation loop for any programming task. Guides the user through defining goals, measurable metrics, and scope constraints, then runs an autonomous loop of code changes, testing, measuring, and keeping/discarding results. Inspired by Karpathy's autoresearch. USE FOR: autonomous improvement, iterative optimization, experiment loop, auto research, performance tuning, automated experimentation, hill climbing, try things automatically, optimize code, run experiments, autonomous coding loop. DO NOT USE FOR: one-shot tasks, simple bug fixes, code review, or tasks without a
How do I install the Autoresearch skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add a-tokyo/aiworkspace --skill autoresearch. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Autoresearch popular?
It lives in a-tokyo/aiworkspace, which has 23 GitHub stars and 5 forks.
Related skills
Production Grade
23a-tokyo/aiworkspace
Principle-engineering posture for production-grade code: reads the repo first, plans before code, matches conventions, pulls latest docs over training recall, and ships the simplest correct change that holds the bar — proper algorithms and data structures, idempotent writes, schema+queries+indexes as one artefact, typed errors, tests in the same diff. Substrate-agnostic; defers to peer skills on their lanes. Use for non-trivial planning, design, implementation, review, or refactoring; RCA and debugging; performance and optimization work; changes touching a database schema, security, infrastruc
Scaffold Skill
23a-tokyo/aiworkspace
Scaffold new AI agent skills with proper symlinking across all IDE tools (Cursor, Claude Code, Codex, etc.). Use when creating a new skill, adding a local skill, or when the user wants to write a custom skill -- either workspace-wide or for a specific project.
Tribunal
23a-tokyo/aiworkspace
Runs a doer -> verifier-panel -> consensus loop to verify a deliverable before it ships. An orchestrator freezes acceptance criteria before implementation, dispatches a doer, then convenes a context-walled panel of independent verifiers - including an adversary with an explicit must-oppose mandate - for evidence-anchored review adjudicated to a SHIP / SHIP_WITH_CAVEATS / ITERATE / BLOCK / ESCALATE verdict logged to a ledger. Use for multi-agent verification of any artifact - code slices, plans, documents, audits - whenever asked to verify a deliverable, vet a plan, run a consensus review or in
Bark Notify
16caopulan/Notification-Skill
Send Bark (day.app) push notifications after Codex completes a task. Use when one Codex or Claude run is finished, or when you need to notify on task completion with device name, project name, status, and summary.
Alternative Uses Task Designer
43NeuroAIHub/awesome_cognitive_and_neuroscience_skills
Domain-validated guidance for designing Alternative Uses Task (AUT) experiments measuring divergent thinking, with parameters for AI-augmented and traditional conditions
Divergent Thinking Scoring
43NeuroAIHub/awesome_cognitive_and_neuroscience_skills
Domain-validated multi-dimensional scoring system for divergent thinking tasks, including fluency, flexibility, originality, and automated semantic distance methods
Browse the most popular Claude skills, all productivity & notes skills, or write your own with the skills generator.