Awesome AGI Agents

Awesome-AGI-Agents

Agents (Agents) Featured resource collection, continuously updated

Articles and videos

name	Introduction	Remark
AI Agents exploded: Software 2.0 is emerging, the next step for OpenAI	Lilian Weng's personal blog post, Lilian is now the Head of Safety Systems of OpenAI and has previously led the Applied AI team of OpenAI. AI Agent is considered the next direction for OpenAI to make efforts. Andrej Karpathy, co-founder of OpenAI, mentioned in a recent public event that "compared with model training methods, OpenAI is currently paying more attention to changes in the Agent field. Whenever a new AI Agents paper comes out, it will be very excited and seriously discussed internally." Earlier, Andrej also commented that AutoGPT is the next stage of exploration of Prompt Engineering.	Original English
"Summary: Agent Driven by the New Big Language Model" - 45,000 words to explain the latest Agent Survey of Fudan NLP and Mihayou	A translated version of the paper "TITLE: The Rise and Potential of Large Language Model Based Agents: A Survey" published by the Fudan NLP team and Mihayou. The author also simplified some of the content and summarized it.	Original text of the paper
AI Agent’s $100 billion question: How to reconstruct a 1 billion knowledge job and start a revolution in software production?	The author believes that the characteristic that an agent product needs to have is to provide product designers and users with room for intervention. There are two most representative categories in practice: the Agent Framework in the middle layer and the Vertical Agent in the vertical field.	-
LangChain Agents - Joining Tools and Chains with Decisions	LangChain Agents--Combining tools and task chains with decision-making	English Youtube video, official introduction to preset agents by LangChain project.

paper

name	Introduction	Remark
LLM-Agent-Survey	LLM-based Autonomous Agents Overview
LLM-Agent-Paper-List	"The Rise and Potential of Large Language Model Based Agents: A Survey"
LLMAgentPapers	LLM-based Autonomous Agents Overview
Toolformer: Language Models Can Teach Themselves to Use Tools	Large language models can learn to use tools	-
HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face	Using ChatGPT as the controller, connect various AI models in the HuggingFace community to complete multimodal complex tasks. The whole process only needs to do is: output your needs in natural language.	Zhihu Chinese Discussion
ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs	The researchers designed a Benchmark (benchmark) that evaluates the ability of LLM to use tools, and a framework for data construction, model training, and evaluation of this scenario - ToolLLM.	[Zhihu Article], [GitHub Open Source Address]
AgentBench: Evaluating LLMs as Agents	AgentBench is the first benchmark to evaluate LLM as agents in a variety of different environments. It contains 8 different environments that can more comprehensively evaluate the ability of LLM to operate as an autonomous agent in various scenarios.	GitHub open source address

Frontier Projects

name	Stars	Introduction	Remark
Auto-GPT		An experimental open-source attempt to make GPT-4 fully autonomous.	The famous AutoGPT project, one of the early attempts of AI Agents
gpt-engineer		Specify what you want it to build, the AI ​​asks for clarification, and then builds it.	Full stack development agent, write the entire project code in GPT!
AgentGPT		Assemble, configure, and deploy autonomous AI Agents in your browser.	Deploy running AI Agents in your browser.
MetaGPT		? The Multi-Agent Framework: Given one line Requirement, return PRD, Design, Tasks, Repo.	MetaGPT, multi-agent framework, enter a sentence of the boss's needs, output user stories/competitive product analysis/Requirements/Data structures/APIs/files, etc.
llama-hub,shopify agent		A customer support agent ? that can interface with @Shopify's ENTIRE GraphQL API Spec (>50k lines!).	The customer support agent built by llama-hub can interact with @Shopify's entire GraphQL API specification (>50,000 lines!)
Agent-LLM		An Artificial Intelligence Automation Platform. AI Instruction management from various providers, has an adaptive memory, and a versatile plugin system with many commands including web browser.	Artificial intelligence automation platform. https://agent-llm.com/
skyagi		SkyAGI implements the idea of ​​Generative Agents and delivers a role-playing game that creates a very interesting user experience.	SkyAGI implements the concept of "general agent" and designs a role-playing game to create a very interesting user experience.
generative_agents		Generative Agents: Interactive Simulacra of Human Behavior.	Stanford and Google researchers created a town of AI agents inspired by the Sims game; the researchers added 25 Generative Agents to the simulated town, which are controlled by ChatGPT and custom code to live independently with highly realistic behavior. Supported by ChatGPT, everyone has their own unique identities, memories, and behaviors and can interact independently, but none of them realize they are living in simulations. Introduction to Chinese
Developer		the first library to let you embed a developer agent in your own app!	The engineer agent gives it a product specification, builds a complete code base for you, provides basic modules, and allows you to have an intelligent developer within your application.

Agents Development Platform

name	Stars	Introduction	Remark
langchain		Building applications with LLMs through composability	ChatGPT application developed to build agents based on LLM. CSV Agent JSON Agent, OpenAPI Agent, Pandas Dataframe Agent, Python Agent, SQL Database Agent, Vectorstore Agent
AutoChain		AutoChain: Build lightweight, extended, and testable LLM Agents.	AutoChain: Build lightweight, scalable and testable LLM agents.
SuperAGI		<⚡️> SuperAGI - A dev-first open source autonomous AI agent framework. Enabling developers to build, manage & run useful autonomous agents quickly and reliably.	The official website builds, manages and runs AI Agents.
superagent		Superagent - Build, deploy, and manage LLM-powered agents.	Official website developers can more easily build, manage and deploy agents into production, including built-in memory, searching documents through vector databases, powerful tools, network hooks, cron tasks and other functions.
ai-town		A MIT-licensed, deployable starter kit for building and customizing your own version of AI town - a virtual town where AI characters live, chat and socialize.	Used to build and customize your own version of the AI ​​town – a virtual town where AI characters live, chat and socialize.
agent-protocol		Common interface for interacting with AI agents. The protocol is tech stack agnostic - you can use it with any framework for building agents.	The "agent communication protocol" that is used by well-known projects such as AutoGPT and smol developer - a common interface for communicating with agents.
autogen		AutoGen is a framework that enables development of LLM applications using multiple agents that can converse with each other to solve task.	AutoGen is a framework that allows the use of multiple agents to develop LLM applications that can talk to each other to solve tasks.
agents		An Open-source Framework for Autonomous Language Agents.	Agents is an open source library/framework for building autonomous language agents.
bisheng		Bisheng is an open LLM devops platform for next generation AI applications.	A leading open source large-scale application development platform, empowering and accelerating the implementation of large-scale application development, helping users enter the next generation of application development mode with the best experience.
Agently		A fast way to build LLM Agent based Application.	For application developers: Agently provides the ability to easily generate powerful Agent instances quickly, allowing developers to easily combine these instances with their own business code.

Curated List

name	Stars	Introduction	Remark
awesome-ai-agents		A list of AI autonomous agents.	Featured resources for agents based on LLM.

Star History