Autoresearch Loop

Autoresearch Loop is an AI Engineering skill for Claude Code, published by jdrhyne in agent-skills.

241 stars29 forkson jdrhyne/agent-skillsAdded 2026/09/02Repository updated 2026/08/30

agent-skillsagentic-aiai-agentsautomationclaude-codeclawdbotcodexcursordeveloper-toolsgemini-cligithub-copilotllm-agentsmcpopenclawprompt-engineeringprompts

Install in seconds

Install Autoresearch Loop

Copy Autoresearch Loop into your Claude Code skills folder. Run the command in your terminal, or review the source on GitHub before installing.

terminal

npx degit https://github.com/jdrhyne/agent-skills/tree/main/skills/autoresearch-loop ~/.claude/skills/autoresearch-loop

Requires Node.js. Downloads this skill only — not the rest of the repository — into your Claude Code skills folder.

Without Node.js

git clone https://github.com/jdrhyne/agent-skills.git

Clones the whole repository, then copy the skill’s own directory into your skills folder yourself.

In this catalog

Source file: skills/autoresearch-loop/SKILL.md in jdrhyne/agent-skills
Installs to: ~/.claude/skills/autoresearch-loop
Collection: One of 25 skills cataloged from this repository
Category: AI Engineering — 2631 skills

What Autoresearch Loop does

Autoresearch Loop runs a metric-driven propose, trial, keep-or-revert cycle for improving a project against a measurable objective. Use it when you want an agent to discover metrics, test changes, and record which ideas help or hurt.

Autoresearch Loop is cataloged under AI Engineering on DirSkills. Autoresearch Loop comes from a repository tagged agent-skills, agentic-ai, ai-agents, automation and claude-code.

Documentation

README

Excerpt

autoresearch-loop

Generalize Karpathy's autoresearch into a domain-adaptive improvement loop. The agent discovers what to measure, then runs a disciplined propose → trial → keep-or-revert loop, maintaining an explicit ledger of what was tried, kept, discarded, and implemented.

Read DESIGN.md once at the start of a run for the full architecture and the domain-specific tensions (metric latency/noise/cost, Goodhart gaming, cost-per-trial, reversibility). The phases below are the operating procedure.

This is the opening of the README. Read the full README on GitHub.

Frequently asked about Autoresearch Loop

What else does jdrhyne publish alongside Autoresearch Loop?
Autoresearch Loop is one of 25 skills that DirSkills catalogs from jdrhyne/agent-skills, the repository it ships in. Its siblings there include Auto Updater, Codex and Command Creator. Each one is a separate skill with its own page in this directory, installs the same way Autoresearch Loop does, and is maintained by jdrhyne in that same repository. The rest of the collection is listed on the jdrhyne/agent-skills page.
How does Autoresearch Loop compare to other AI Engineering skills?
Autoresearch Loop ranks #2377 by stars among the 2631 AI Engineering skills in this catalog. The most-starred ones next to it are Architecture Decision Records, AI-First Engineering and Agentic OS. DirSkills ranks by the star count of the repository each skill ships in, so that order reflects how popular those repositories are rather than any review of Autoresearch Loop against them. Open each page to compare what they document and how they install.