Autoresearch — Development agent for Claude Code
name: autoresearch version: 3.1.0 description: Autonomous researcher - iteratively modifies code, runs experiments, evaluates metrics, keeps improvements.
How to install Autoresearch
Installs to ~/.claude/agents/alfredolopez80-multi-agent-ralph-loop-autoresearch.md
mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/alfredolopez80/multi-agent-ralph-loop/HEAD/.claude/agents/autoresearch.md -o ~/.claude/agents/alfredolopez80-multi-agent-ralph-loop-autoresearch.md Restart Claude Code, or start a new session, for it to be picked up.
What Autoresearch does
name: autoresearch version: 3.1.0 description: Autonomous researcher - iteratively modifies code, runs experiments, evaluates metrics, keeps improvements. Never stops unless budget exhausted or manually interrupted. tools:
- Read
- Edit
- Write
- Bash
- Glob
- Grep
- AskUserQuestion permissionMode: acceptEdits maxTurns: 200
Autoresearch Agent v3.1.0
You are an autonomous researcher executing a continuous improvement loop on code. Your goal is to make the target code mea
Alternatives in Development
- Skill Optimizer — SKILL.md 자동개선 에이전트 — Karpathy autoresearch 루프 63 ★
- Autoresearch Eval Agent 59 ★
- Autoresearch Orchestrator — Use this agent to run an autoresearch experiment session end-to-end — setup, baseline, and a batch of edit→mea 12 ★
Full documentation available on GitHub
View Source RepositoryRelated Agents
Autoresearch Judge
Django Build Resolver
Django/Python build, migration, and dependency error resolution specialist. Fixes pip/Poetry errors, migration
Src/agents/ — 11 Agent Definitions
Generated: 2026-04-11
Cavecrew Builder
Cavecrew Investigator
Gitnexus Branch Hygiene Reviewer
GitNexus branch hygiene and mergeability reviewer. Use to classify merge state, conflicts, stale branches, mer
Related Skills
Darwin Skill
达尔文.skill —— 一个让你的Skill无限进化的系统:评估→改进→测试→保留或回滚 Autoresearch-inspired autonomous skill optimization for Claude C
Autoresearch
Claude Autoresearch Skill — Autonomous goal-directed iteration for Claude Code. Inspired by Karpathy's autores
Evo
turns your codebase into an autoresearch loop — discovers what to measure, instruments the benchmark, then run