A list of Agentic Frameworks and Tools

I've collected the following list of Agentic Frameworks and Tools over time. Instead of them occupying 8 749 browser tabs, I made this list to I can find it myself again and hopefully someone else finds it useful.

GitHub Repos

** Skyvern ** [https://github.com/Skyvern-AI/skyvern]

• Their Description [https://www.skyvern.com] - Traditional approaches to browser automations required writing custom scripts for websites, often relying on DOM parsing and XPath-based interactions which would break whenever the website layouts changed.

Instead of only relying on code-defined XPath interactions, Skyvern relies on prompts in addition to computer vision and LLMs to parse items in the viewport in real-time, create a plan for interaction and interact with them.

• My Take - I am keen to test this one, have not tried it myself. Given their commercial / cloud offering, I suspect the repo by itself would be crippled to incentivise user to the commercial offering.

** OpenAI Swarm ** [https://github.com/openai/swarm]

• Their Description - The primary goal of Swarm is to showcase the handoff & routines patterns explored in the Orchestrating Agents: Handoffs & Routines cookbook. It is not meant as a standalone library, and is primarily for educational purposes.

• My Take - I've started building some example workflows with swarm and it's very easy to understand and implement. The resulting structure does need some active management and the framework itself only provides the minimum capabilities to deplyment at scale would not be possible as you lack the reporting and analytics components. But it's awesome to prove a concept.

AutoGPT [https://github.com/Significant-Gravitas/AutoGPT]

• Their Description - AutoGPT is a powerful platform that allows you to create, deploy, and manage continuous AI agents that automate complex workflows.

• My Take - The project still seems relatively active, although activity has diminished since November 2023. 167 000 Stars on GitHub shows that it did get a lot of attention. This one is definately high on my to-do list to test. They have a gloud-hosted Beta version, so there is a commercial offering in the works.

ChatDev [https://github.com/OpenBMB/ChatDev]

• Their Description - ChatDev stands as a virtual software company that operates through various intelligent agents holding different roles, including Chief Executive Officer , Chief Product Officer , Chief Technology Officer , programmer , reviewer , tester , art designer . These agents form a multi-agent organizational structure and are united by a mission to "revolutionize the digital world through programming." The agents within ChatDev collaborate by participating in specialized functional seminars, including tasks such as designing, coding, testing, and documenting

• My Take - This started out quite promising, but the repo seems to have become quiet. I suspect the reasoning capability offered by something like GPT-o1 when combined with a design / workflow as implemented by Replit Agents will make this somewhat obsolete.

• Some more reading - Building AI Agent Workforce with MetaGPT & ChatDev [https://aimation-ed.medium.com/building-ai-agent-workforce-with-metagpt-chatdev-4a1c80506ddb]

• And the associated Video [https://www.youtube.com/watch?v=pJwR5pv0_gs]

CrewAI [https://github.com/crewAIInc/crewAI]

• Their Description - CrewAI is designed to enable AI agents to assume roles, share goals, and operate in a cohesive unit - much like a well-oiled crew. Whether you're building a smart assistant platform, an automated customer service ensemble, or a multi-agent research team, CrewAI provides the backbone for sophisticated multi-agent interactions.

• My Take - Similarly, I think CrewAI is useful in the sense that it solves more complex problems by allowing smaller, less specialised models, to tacke sections of a problem. More powerful models like GPT-o1 will probably be able to achieve the same outcome withouth the orchastration. But if you wanted to run these workflows provately, then CrewAI is an option where GPT-o1 is not.

• Paid-for Offering [https://www.crewai.com/ ]

MetaGPT [https://github.com/geekan/MetaGPT]

• Their Description - Software Company as Multi-Agent System. MetaGPT takes a one line requirement as input and outputs user stories / competitive analysis / requirements / data structures / APIs / documents, etc. Internally, MetaGPT includes product managers / architects / project managers / engineers. It provides the entire process of a software company along with carefully orchestrated SOPs.

• My Take - Looks like an interesting concept. Some activity on GitHub but seems to have gone very quiet since April 2024. Not as popular as some of the others.

• Some more reading - Building AI Agent Workforce with MetaGPT & ChatDev [https://aimation-ed.medium.com/building-ai-agent-workforce-with-metagpt-chatdev-4a1c80506ddb]

• And the associated Video [https://www.youtube.com/watch?v=pJwR5pv0_gs]

AgentVerse [https://github.com/OpenBMB/AgentVerse]

• Their Description - AgentVerse is designed to facilitate the deployment of multiple LLM-based agents in various applications. AgentVerse primarily provides two frameworks: task-solving and simulation. Task-solving: This framework assembles multiple agents as an automatic multi-agent system (AgentVerse-Tasksolving, Multi-agent as system) to collaboratively accomplish the corresponding tasks. Applications: software development system, consulting system, etc. Simulation: This framework allows users to set up custom environments to observe behaviors among, or interact with, multiple agents.

• My Take - Project seems to have gone stale.

XAgent [https://github.com/OpenBMB/XAgent]

• Their Description - XAgent is an open-source experimental Large Language Model (LLM) driven autonomous agent that can automatically solve various tasks. It is designed to be a general-purpose agent that can be applied to a wide range of tasks. XAgent is still in its early stages, and we are working hard to improve it. Our goal is to create a super-intelligent agent that can solve any given task!

• My Take - This project is essentially dead. No activity since Jan 2024.

ToolBench [https://github.com/OpenBMB/ToolBench]

• Their Description - This project (ToolLLM) aims to construct open-source, large-scale, high-quality instruction tuning SFT data to facilitate the construction of powerful LLMs with general tool-use capability. We aim to empower open-source LLMs to master thousands of diverse real-world APIs. We achieve this by collecting a high-quality instruction-tuning dataset. It is constructed automatically using the latest ChatGPT (gpt-3.5-turbo-16k), which is upgraded with enhanced function call capabilities. We provide the dataset, the corresponding training and evaluation scripts, and a capable model ToolLLaMA fine-tuned on ToolBench.

• My Take - This project has gone very quiet since November 2023.

Camel AI [https://github.com/camel-ai/camel]

• Their Description - The rapid advancement of conversational and chat-based language models has led to remarkable progress in complex task-solving. However, their success heavily relies on human input to guide the conversation, which can be challenging and time-consuming. This paper explores the potential of building scalable techniques to facilitate autonomous cooperation among communicative agents and provide insight into their "cognitive" processes. To address the challenges of achieving autonomous cooperation, we propose a novel communicative agent framework named role-playing.

• My Take - Very active project with lots of activity. Have not had the time to test it. Similar to some other project, trying to "Solve" the Though part. Could be a useful way to get gpt-o1 level results using smaller llama3.2-type models.

Academic Research / Papers

• Cognitive Architectures for Language Agents [https://arxiv.org/abs/2309.02427?utm_source=div.beehiiv.com&utm_medium=referral&utm_campaign=advanced-rag-series-generation-and-evaluation]

• Self-RAG: Learning to Retrieve, Generate and Critique through Self-Reflection [https://arxiv.org/abs/2310.11511?utm_source=div.beehiiv.com&utm_medium=referral&utm_campaign=advanced-rag-series-generation-and-evaluation]

• A Path Towards Autonomous Machine Intelligence [https://openreview.net/pdf?id=BZ5a1r-kVsf]

• Abjective-Driven AI (Meta and NY University) [https://drive.google.com/file/d/1Ymx_LCVzy7vZXalrVHPXjX9qbpd9k_bo/view?fbclid=IwAR0Qzn4wLWha1Vk_ivjeyfJcBWDVfnt1uQxflOMhDFXvlYcmC9AQl8xLffE_aem_AUH59ZzK4cl-ycv1f4q_SBNLz3-2OPhEQFnRSydhIOyVX9qmqVOll3zCOJfc12JGreZ6f3odzruHDa2TprNvkEZW]