autoresearch
An AI agent for automatically iterating and improving code, supporting Claude Code, OpenCode, and Codex.
- Region
- Overseas
- Pricing
- Free
- Open source
- Yes
- GitHub Stars
- ★ 4.8k
- Source
- GitHub
- Added
- 2026-06-06
- Last verified
- 2026-06-06
Overview
autoresearch is an AI agent that automatically iterates to improve code, supporting Claude Code, OpenCode, and Codex. It is based on Karpathy's autoresearch concept, achieving continuous code optimization through defined goals, metrics, and iterative cycles. Users need only set the target, and the agent will autonomously run loops until the desired outcome is achieved.
Key features
- ▪Automatic iteration and code improvement
- ▪Supports Claude Code, OpenCode, and Codex
- ▪Based on Karpathy's autoresearch philosophy
- ▪13 commands, 9 safety hooks
Use cases
Pros
- ＋Reduces manual intervention
- ＋Improves code quality
- ＋Saves development time
Limitations / notes
- －Requires clear goals and measurable criteria
Who it's for
This overview was compiled by AI from public sources and may contain inaccuracies — please refer to the official site.
FAQ
Which programming tools does autoresearch support?
Supports Claude Code, OpenCode, and OpenAI Codex.
How do I get started with autoresearch?
Set your goal and metrics, then start the agent for automatic iteration.
Similar agents
Cursor
FeaturedAI-powered code editor with Agent mode
Dify
FeaturedOpen-source LLMOps + Agent orchestration platform
扣子 Coze
FeaturedByteDance's no-code Agent/Bot building platform
Trae
FeaturedAI IDE (with Agent) from ByteDance
LangChain
FeaturedThe most mainstream LLM/Agent orchestration framework
Devin
FeaturedAutonomous software engineer Agent by Cognition
Something wrong? Let us know on the About page and we'll fix it.