OSS Alternative - Discover Top Open Source Alternatives to Popular Software

wanshuiyin/Auto-claude-code-research-in-sleep

Automates machine learning research workflows using LLM agents, enabling autonomous idea discovery, experiment execution, and paper review with a lightweight, framework-agnostic approach.

Core Features

Autonomous ML Research: Automates idea discovery, experiment execution, and paper review.

Cross-Model Collaboration: Orchestrates multiple LLMs (e.g., Claude Code, Codex) for diverse roles like research and critical review.

Lightweight & Framework-Agnostic: Built with plain Markdown files, zero dependencies, no lock-in to specific frameworks or databases.

Persistent Knowledge Base: Includes a "Research Wiki" for persistent storage of papers, ideas, experiments, and claims.

Self-Evolution Capabilities: Features "Meta-Optimize" for analyzing logs and proposing skill improvements.

Detailed Introduction

ARIS (Auto-Research-In-Sleep) is an innovative project designed to revolutionize machine learning research through autonomous AI agents. It provides a lightweight, Markdown-only skill-based workflow that enables LLMs like Claude Code, Codex, and others to perform complex research tasks, including idea generation, experiment automation, and critical review. By fostering cross-model collaboration and offering a framework-agnostic approach, ARIS minimizes setup overhead and maximizes flexibility, allowing researchers to automate their work and accelerate discovery without vendor lock-in. Its unique features like a persistent Research Wiki and self-evolution capabilities further enhance its utility for advanced AI-driven research.