AI Agents

Adala is an Autonomous Data (Labeling) Agent framework designed for reliable and controllable data-driven agents. It excels in data labeling and processing tasks, supports autonomous learning via reflection, and features a flexible and extensible runtime.

A recommender system simulator featuring 1,000 LLM-based agents initialized from the MovieLens-1M dataset, each simulating human-like interactions with movie recommendation systems.

Low-code framework for building and testing autonomous agents and cognitive architectures compatible with multiple LLMs like OpenAI, Claude, and Oobabooga. Supports agent memory and database-agnostic configuration.

Browser-based no-code platform allowing users to deploy GPT agents by specifying goals, which the agent then plans and executes autonomously. Built with OpenAI functions and supports databases like Pinecone and pg_vector.

Desktop app for building, managing, and chatting with agents. Integrates with Open Interpreter and MemGPT. Offers group chat and GUI-based interaction.

Agents is a comprehensive framework for building LLM-based agents with symbolic control, multi-agent communication, and integration with tools and standard operating procedures.

AgentVerse is a platform to simulate and coordinate multiple agents solving tasks within customizable environments, offering flexibility for collaboration and emergent behavior exploration.

AI Legion is an LLM-driven autonomous agent framework that emphasizes collaborative task execution between agents. Console-based DM interface.

Aider is a command line tool that lets you pair program with GPT-3.5/GPT-4, to edit code stored in your local git repository. It works well with pre-existing, larger codebases and ensures edits from you and GPT are committed to git with sensible commit messages.

An agent in the form of a chatbot that independently plans tasks given in natural language and dynamically creates an agents-calling tree to execute tasks. Features interaction between agents for fault tolerance and automatic expansion.

A framework by Microsoft for developing LLM applications around multi-agent conversations. AutoGen simplifies workflows and supports human-agent collaboration, customizable patterns, and LLM-optimized interactions.

An experimental open-source project that chains together LLM thoughts and actions to achieve user-defined goals. Offers plugins, internet browsing, file storage, and various execution capabilities.

Automata provides autonomous code generation based on project context. It aims to become a fully self-programming system using SymbolRank algorithm and modular LLM integration.

AutoPR automatically generates PRs by planning fixes, writing code, creating branches, and opening pull requests based on labeled GitHub issues.

A GPT-3.5 based HR assistant capable of answering HR-related queries using integrated internal tools like policies, employee data, and calculators.

A task-driven autonomous agent framework using OpenAI GPT-4, LangChain, and Pinecone. It manages a task list from a main objective and generates subtasks adaptively.

An enhancement of BabyAGI with a more complex task management prompt enabling expanded analysis, synthesis, and integrated function handling.

A simplified mod of BabyAGI with basic agents, reusable tools, and a lightweight 300-line implementation for rapid experimentation.

BabyDeerAGI adds parallel task execution, user input integration, and a lightweight 350-line architecture over the original BabyAGI model.

An advanced BabyAGI variant introducing dynamic task lists, skill creation, and self-editing agent capabilities.

An agent designed to test what happens when you combine CLI and LLM, based on BabyAGI. Executes shell commands, automates programming and environment setup, and can handle general tasks like text generation and translation. Recommended to run in a virtual environment.

A mod of BabyElfAGI with self-improving task lists (FOXY method), novel chat UI with parallel tasks, and new skills like DALL-E integration and Airtable search.

A loop-based data exploration and analysis tool for non-programmers. Supports multiple models, maintains memory, builds prompts dynamically, and generates code for analysis and visualization.

An early-stage research project for a general-purpose agent capable of a wide range of tasks. Not intended for production use at this time.

An open-source AI debugging agent for VSCode that uses LLMs to help identify and fix backend code errors. Leverages VSCode API, LSP, and print statement debugging.

A GPT-4 powered semantic code search engine for Rust and Typescript. Features fast code search, regex matching, and natural language queries for local and remote repositories.

An autonomous AI agent with CLI, RESTful/WebSocket API, and integrated tools. Supports OpenAI and Azure models, code interpreter, data query, and Docker-based safe execution.

An AI agent that keeps npm dependencies up-to-date by bumping package versions and generating code fixes for breaking changes. Supports GPT-4-turbo.

An open-source scheduling assistant built on Cal.com. Books meetings, summarizes your week, and finds time with others using natural language. Powered by GPT-4 and LangChain.

An open-source library for the study of autonomous and communicative agents. Features user and assistant agents, and a community for agent research.

An AI assistant designed to let you interactively query your codebase using natural language. Utilizes vector embeddings, chunking, and OpenAI's language models to help you navigate your code efficiently.

An AI code assistant from Sourcegraph that writes code and answers questions by reading your entire codebase and the code graph.

An open-source autopilot for software development—brings the power of ChatGPT to VS Code. Features include answering coding questions, editing in natural language, and generating files from scratch.

A framework for orchestrating role-playing, autonomous AI agents. Empowers agents to work together seamlessly, tackling complex tasks. Built on LangChain, with modular design and collaborative decision-making.

A framework for systematically navigating the power of AI to perform end-to-end scientific research, from raw data to comprehensive, human-verifiable research papers. Guides LLM and rule-based agents through the scientific process.

A no-code platform for creating AI chatbots trained on your own data. Deploy to website, Slack, Crisp, or Zapier. User-friendly interface for managing datastores and chat with your data.

Leverages LLMs to provide fast and effective demo creation for applications. Automates prototyping, processes prompts to generate relevant applications, and integrates with LangChain.

A team of virtual developers that can build microservices from your instructions. Includes virtual Product Manager, Developer, and DevOps for concept to deployment.

An Agentic AI Software Engineer that understands high-level instructions, breaks them down, researches, and writes code to achieve objectives. Open-source alternative to Devin.

An open-source alternative to Devin by Entropy Research. Designed for autonomous software engineering tasks.

AI-driven software development automation solution that translates natural language requirements into working software. Enhances quality, lowers costs, and improves collaboration.

Agent system enabling deployment across cloud, PC, and mobile. Includes prompt compiler, execution sandboxing, and integration via Python or plain English.

An open-source SDK for building modular, pluggable, and extensible AI agents and frameworks.

Compiler for converting functional markdown specs written in natural language into executable code. Open-sourced by Uilicious.

Persona-adaptive agent platform that changes behavior by selecting appropriate agent personas for a given task in real-time.

Framework for quickly scaling Jupyter-based multi-agent prototypes into production apps. Supports AutoGen and multi-framework deployment.

Low-code open-source tool for visually building LLM flows and agents, ideal for RAG or multi-step LLM applications.

AI developer assistant for Node.js apps. Enables multi-section prompt workflows for fast project generation. Supports GPT-4 models.

AI team assistant for DevOps. Executes tasks across k8s, AWS, Argo, GitHub. Extensible via APIs and Slack integration.

AutoGPT/BabyAGI-inspired browser-based agent executing tasks via interactive UI. Example uses include ordering coffee or analyzing markets.

A multi-agent language game environment for LLMs, providing a framework for building interactive environments and communication/collaboration capabilities for AIs. Includes web UI and CLI.

A virtual software company driven by intelligent agents in different roles (CEO, CTO, programmer, etc.) collaborating to design, code, test, and document software. Modular, extensible, and open-source.

LangChain agent for chemistry-related tasks, integrating 13 expert-designed tools to augment LLM performance in chemistry. Demonstrates effectiveness in automating chemical tasks.

An agent that can plan, write, debug, and test code autonomously. Designed to develop code for or with the user, with feedback-driven improvement.

An intelligent assistant serving the entire software development lifecycle, powered by a Multi-Agent Framework, working with DevOps Toolkits, Code&Doc Repo RAG, etc.

A robust, all-in-one GPT interface for Discord, supporting multi-modality, code execution, advanced data analysis, and more.

An AI agent that generates an entire codebase based on a prompt. Designed to be easy to adapt, extend, and learn your code style.

Easily migrate your codebase from one framework or language to another. Supports multiple LLMs, unit test generation, and workflow customization.

An AI agent that codes entire scalable apps from scratch, with developer oversight. Research project for production-ready app generation.

A GPT-based autonomous agent that does comprehensive online research on any topic, producing detailed, factual, and unbiased reports.

An agent that lets you converse with your files, manage AI prompt storybooks, and supports multiple languages and IDEs. All data is local and private.

A graph-based framework for LLM-based agents, enabling the building of agents from graphs and automatic self-organization of agent swarms.

A platform for building, debugging, and deploying collaborative agents and cognitive workflows. Includes a no-code graph editor and LangChain-based runtime.

A system to connect LLMs with the ML community, featuring task planning, model selection, and execution via Hugging Face models.

A lightweight, extensible Python framework for building LLM-powered multi-agent applications, inspired by the Actor Framework.

A Plan-Validate-Solve agent for workflow automation, supporting read/write operations on tools like GitHub, HubSpot, and Airtable.

A minimalistic library for building agents that leverage LLMs to automate tasks through commands and tool integrations.

A no-code platform to build LLM agents, workflows, and applications with your data. Supports all major model providers.

Chat with your documents on your local device using GPT models. 100% private, no data leaves your device.

A modular, extensible re-implementation of AutoGPT as a Python package, with human-in-the-loop and full state serialization.

A multi-agent generation framework for producing large codebases or books from a single prompt, with error correction and memory.

A codebase agent that labels, assigns, and answers questions for new issues and PRs. Can search code, run scripts, and write code.

AIDE for creating, deploying, scaling, and monetizing useful AI agents and prompt chaining. Visual development and cognitive architecture.

Open Source Hybrid AI Search Engine for accurate answers from the Internet, bookmarks, notes, and docs. Self-hosted vector DB and embedding.

A system that manages different memory tiers in LLMs to provide extended context. Supports perpetual chatbots and local/SQL data chat.

AI tool that assists with coding tasks from the command line, coordinating edits across multiple files and locations.

A multi-agent framework that returns PRD, design, tasks, or repo from a one-line requirement. Supports collaborative software engineering.

A minimal general-purpose autonomous agent based on GPT-3.5/4. Supports stock analysis, security tests, art, and more.

Implementation of the Multiagent Debate paper, showing improved factuality and reasoning in LLMs through agent debate.

Experimental multi-agent system where multiple expertGPTs collaborate to perform a task, each with their own memory and communication.

Open-source AI for generating unit tests targeting code vulnerabilities, boosting developer productivity and code security.

Natural Language-Based Societies of Mind: a concept and framework for societies and communities of agents communicating in natural language.

Open platform for language agents, with web UI, productive agents, and easy deployment. Fully open-sourced and extensible.

Open-source AGI R&D platform for both benchmark and open-ended tasks. Supports multi-modality, task decomposition, and RL from task feedback.

OpenDevin: Code Less, Make More. Open-source project to replicate, enhance, and innovate beyond the original Devin model.

Open-source interpreter that lets LLMs run code on your computer to complete tasks. Works locally and from a terminal UI.

Development toolkit for prompt management, versioning, publishing, collaboration, troubleshooting, and observability.

A tool for private interaction with documents, built with LangChain, GPT4All, LlamaCpp, Chroma, and SentenceTransformers.

Personalized AI assistant that decomposes problems, offers solutions, and lets you use Agent actions to automate your flows.

Open-source React.js autonomous LLM agent that generates and composes React components from user stories, using atomic design principles.

Lets multimodal models operate a computer, using the same inputs and outputs as a human operator to reach objectives.