This repository is an exhaustive, community-driven collection of resources on LLM-powered agents, targeting researchers and developers. It aims to catalog papers, open-source projects, and discussions related to agents capable of autonomous task solving, multi-agent collaboration, and complex reasoning, providing a valuable overview of the rapidly evolving field.

How It Works

The project functions as a curated list, categorizing and linking to a wide array of academic papers, code repositories, and blog posts. It covers diverse agent capabilities such as planning, tool usage, multi-agent cooperation, web navigation, robotics, and gaming. The collection is updated regularly, with contributions welcomed via pull requests and issues.

Quick Start & Requirements

This is a curated list, not a software project. No installation or execution is required.

Highlighted Details

• Extensive categorization of papers and projects across numerous agent application domains.
• Includes papers with high citation counts (🔥) and those accepted by reputable conferences/journals (📖).
• Features links to code repositories, project pages, and blogs for deeper exploration.
• Covers foundational frameworks like AutoGen, LangChain, and MetaGPT.

Maintenance & Community

The repository is noted as "not under active maintenance" but welcomes contributions via PRs and issues. It lists numerous contributors and acknowledges Yupeng Hou as the initial builder and maintainer.

Licensing & Compatibility

The repository itself is a collection of links and does not have a specific software license. Individual linked projects and papers will have their own licenses.

Limitations & Caveats

The collection primarily contains papers published before October 2023, with some later additions. The repository is not actively maintained, meaning new developments might not be immediately incorporated.