webfuse-com/awesome-autoresearch

What it solves

This project addresses the challenge of discovering and organizing autonomous AI research systems. It serves as a central index for tools that implement "autoresearch"—a concept where AI agents run iterative loops to improve code, optimize metrics, or conduct scientific research with minimal human intervention.

How it works

It acts as a curated directory of open-source projects inspired by the original karpathy/autoresearch framework. These projects typically involve an agentic loop that defines a goal, executes experiments, evaluates the results, and uses that feedback to refine its approach. The index categorizes these tools by their specific focus, such as general-purpose optimization, scientific discovery, platform-specific adaptations, and evaluation frameworks.

Who it’s for

This resource is for developers, AI researchers, and enthusiasts who want to build or experiment with autonomous agent systems. It is particularly useful for those looking for existing frameworks to automate tasks like code optimization, machine learning model improvement, or end-to-end scientific paper writing.

Highlights

• A comprehensive collection of general-purpose autonomous improvement loops.
• Specialized research-agent systems for end-to-end scientific discovery (from literature review to paper drafting).
• Hardware-specific ports for running research agents on Apple Silicon, Windows, and other platforms.
• Links to evaluation benchmarks and notable case studies for measuring agent performance.
• A community-driven list that welcomes contributions to keep the index up to date with the latest agentic research.