Luo Baishun

MGX - The first AI dev team

MGX (MetaGPT X) is a multi-agent AI platform based on real software SOPs. Chat with the AI team leader, product manager, architect, engineer, and data analyst 24/7 to create websites, blogs, shops, analytics, games, or anything else you want to try.

Add a comment

Replies

Best
Luo Baishun

MGX reminds me of Devin, but with a unique twist—five specialized agents working in real-time to handle data, coding, and design. I haven’t seen anything like this before! 

@luobaishun Totally agree! The 5 agents are so cuuuute!!

Alex

@luobaishun Thank you for sharing your thoughts!

I'd like to share the design inspiration behind MGX. Before creating MetaGPT (the open-source project behind MGX), I was a software company CEO who coded for eighteen years, using up to 21 programming languages at my peak. I understand what it takes to produce effective code—it requires a comprehensive set of SOPs that are challenging to implement in practice. Getting humans to adopt these procedures and play specific roles effectively consumed all my energy.


During my years managing the company, I gradually came to understand why software companies need product managers, architects, engineers, QA specialists, and project managers, and why these roles form a more effective structure together. In essence, we're shaping agent societies based on human social structures, though I often wonder if intelligent agents might be more like the Borg from Star Trek—but who knows? (I almost renamed the open-source project to "Borg"!)


I'm eager to share my complete perspective with everyone, so I'll be writing a blog post that will be featured on the mgx.dev homepage. I welcome everyone to follow along and continue this conversation.

Manu Hortet

This looks really interesting, would love to try! I'm very bullish on multi-agent systems. Congrats on the launch and good luck! 👏

How does it compare with current tools like Devin, or IDEs like Cursor?

Mike

@manuhortet Thank you so much for your enthusiasm! 🙏 We're also excited about the potential of multi-agent systems. Would love to hear about your experience once you give it a try. Feel free to join our Discord community (https://discord.gg/ZRHeExS6xv) to share your journey! 🚀

Alex

@manuhortet Thanks for asking about the comparison!

MGX takes a fundamentally different approach as a multi-agent system where specialized AI agents (Product Manager, Architect, Data Analyst, Engineer, etc.) collaborate like a real development team. While Devin operates as a single AI developer and Cursor enhances the IDE experience, MGX recreates the entire software development lifecycle with agents that have different expertise and goals - just like human teams.

This multi-agent architecture gives MGX several unique advantages:

  • More specialized expertise for each development phase

  • Reduced context pollution between different stages of development

  • Lower computational costs since agents only focus on their specific tasks

  • The ability to swap or upgrade individual agents independently

  • A more intuitive development process that mirrors how human teams actually work

Currently, MGX excels at web development with various frontend templates, Supabase backend integration, and data visualization capabilities. The system maintains context throughout iterations, allowing projects to continuously evolve.

Each tool has its strengths depending on your needs. What kind of development work are you looking to do? 🚀

Bob

@manuhortet Thank you so much for your interest and kind words! 🙏

As a core developer, I'm excited to share what makes MetaGPT X unique. While Devin and Cursor are excellent tools, our approach is fundamentally different. We're built on the open-source MetaGPT project, which has been focusing on multi-agent collaboration since 2023.

Our key differentiation is our deep commitment to advancing core multi-agent technology. Rather than just creating a coding assistant, we've developed a sophisticated system where multiple specialized agents collaborate like a real software team - from product managers to architects to developers. Each agent has its own expertise and role, working together to solve complex problems.

I'd love to hear your thoughts on multi-agent systems! What aspects of agent collaboration interest you the most? We're always eager to discuss and explore new possibilities in this space.

David
Maker

@manuhortet Thanks so much! 🚀 We're also really excited about the potential of multi-agent systems. As for the comparison, MGX aims to bring more flexibility and dynamic learning to the table, with a focus on streamlining development through natural language interaction. While tools like Devin and Cursor are interesting, MGX is designed around a more collaborative paradigm, where multiple agents work together to tackle development challenges. We're still evolving, but we believe the innovation in this space is huge!👏

Jennie Weng
Congrats on the launch! Wondering, using this tool how long does it take to build up a brand new software? What kind of complexity of a software MGX can support? Thank you!
Mike

@jennieweng Thank you for your kind words and great questions! 🙌

The development timeline really depends on the project scope and complexity. While simpler projects like personal websites or data dashboards can be completed in hours to days, more complex applications might take longer. But here's the thing - the AI team works 24/7, and the quality of the output is worth the wait!

As for complexity, I'd love to show you what MGX can do! Check out our AppWorld (MGX) where you can see live examples. Currently, we're focusing on web development scenarios including personal portfolios, blogs, websites, online stores, aggregator pages, and subscription pages. For backend development, we support integration with Supabase. Additionally, MGX shows excellent performance in data analysis and data science-related scenarios.

Here are some cool examples that showcase our capabilities:

What kind of project do you have in mind? Would love to hear your thoughts and discuss how MGX could help bring it to life! 💡

Alex

@jennieweng I can share a personal example: Over a decade ago, I tried creating a personal blog. I spent months learning WordPress, Linode, PHP, JavaScript, and struggling to fix WordPress bugs while making themes responsive. After all that effort, I only managed to create a mediocre responsive page, which made me doubt my intelligence and question whether I was suited to be a product manager or UI designer.


Fast forward to today - I used MGX for just one hour and the AI agents were able to analyze my needs and create a complete personal website (https://chenglin-wu-card-preview-0i12md.mgx.dev/) without me writing a single line of code.

MGX currently handles web applications of moderate complexity well, with particular strengths in:

  • Frontend development with various templates and frameworks

  • Backend integration with Supabase

  • Data visualization capabilities

  • Full-stack web applications with standard database operations

The multi-agent architecture really shines for projects where you need multiple specialized skills working together - product management, architecture, frontend, backend, etc. What kind of software project are you thinking about building?

Bob

@jennieweng Thank you for your questions!

Based on our testing, MetaGPT X typically takes 15-30 minutes to generate initial architecture and core code for standard web applications.

The actual development time depends on the clarity of requirements and project complexity. Our multi-agent system ensures each component (architecture, code, tests) is developed with rigorous standards.

What kind of project are you looking to build? Would love to provide more specific details based on your use case.

David
Maker

@jennieweng Thanks so much, Jennie! 🚀 Great questions! Based on our testing, for a simple 1-2 page website, it takes around 5-10 minutes to build. For more complex projects, like those with 3-5 pages, multiple page transitions, and various operations or business logic (e.g., handling user authentication with OAuth, implementing role-based access control, managing data processing workflows, integrating APIs for real-time data updates, or implementing multi-step forms and payment gateways), it typically takes about 15-30 minutes.

Currently, MGX is particularly strong with front-end projects, website development, and data analysis. It excels with languages like JavaScript, Vue, Node.js, and Streamlit, and can also handle machine learning tasks, particularly in Python. We're continuously improving its capabilities to support even more sophisticated use cases! 😊

Norvan

👏👏👏ohohoh~~~~~~~~~~Wow!!! 🎉🎉🎉Absolutely amazing!


It far exceeds my expectations — this is truly going to be an incredible product!!!!👍👍👍

Mike

Thank you for your incredible enthusiasm! 🙌 Your energy is contagious! We're working hard to push the boundaries of what's possible with AI development teams. We have some exciting updates coming soon that might surprise you even more! ✨

Bob

@cc_norvan Thank you for your attention and support. You can find more interesting examples on our AppWorld (inside https://mgx.dev/). You're also welcome to join our Discord community (https://discord.gg/ZRHeExS6xv). Looking forward to further communication with you.

David
Maker

@cc_norvan Thank you so much for the incredible feedback! 🙌🎉 We're beyond excited that it’s exceeding your expectations! This kind of support really fuels us to keep pushing forward and making the product even better. Stay tuned—we’re just getting started! 🚀👍 Feel free to explore more MGX features here: MGX Help Center

William Scott

MGX sounds super ambitious — love the concept of turning AI agents into a full dev team that actually follows proper SOPs. Quick question: how much guidance does a user need to provide upfront? Do I have to write out a detailed spec, or can I just throw in a vague idea like “build me a blog for travel stories” and the team figures out the rest?


Also, curious how the AI handles back-and-forth revisions — do the agents keep track of previous changes, or does it feel like starting over each time? Would love to hear if anyone’s already built something cool with MGX!

Mike

@williamrobertscott 
Thanks for your great questions! You've touched on one of our core design principles - making development as effortless as possible!

The beauty of MGX is that you can indeed just share your vision, and let the AI team handle the execution. Simply tell them your goal clearly, and they'll figure out the details, propose features, and handle the implementation.

Here are some examples of what others have built with minimal input:

The best part? You can instantly remix any of these projects and customize them with your content. No need to start from scratch!

Alex

@williamrobertscott That's an excellent question. We have numerous examples in AppWorld where you can click "view chat" to see the prompts used. They vary tremendously - from one-line requests to detailed PRDs.

Humans are remarkably powerful communicators. Saying "Write a Linux" represents tens of millions of lines of code, while "Make a personal page" might mean hundreds. Yet humans can leverage precise language to accomplish either. We view natural language programming as a process of transforming "ambiguous to precise" instructions, maximizing leverage at each step. A decade ago, using C macros might give you a 10x text leverage (one byte controlling ten). Today, we're seeing 1000x or greater.

However, language contains many ambiguities. "Buy an apple" could refer to fruit or electronics, and involves considerations about price, quantity, weight, origin, logistics, amount, channel, brand, timing, quality control, organic status, expiration date, and countless combinations. AI doesn't yet have human-like preferences - it functions more as a probability-driven stateless function.

I encourage thoroughly exploring MGX alongside us to discover its boundaries and help improve it. Its leverage is incredibly high, with efficiency far exceeding humans for shorter code generation (roughly under 5,000 lines). I'll be publishing a blog post specifically addressing my thoughts on this soon.

Regarding revisions - yes, MGX agents do maintain context across iterations, allowing them to track previous changes rather than starting from scratch each time. This enables continuous refinement and development that builds on prior work.

Bob

@williamrobertscott Thank you for your thoughtful questions! As a core developer, I can share our approach to these challenges.

MGX can work with both detailed specs and high-level ideas. For a 'travel blog' request, our PM agent will first generate structured requirements and confirm with you. This ensures alignment while saving your time on detailed specifications.

Regarding iterations, our agents maintain context throughout the development process, tracking changes and design decisions. Each revision builds upon previous work rather than starting fresh.

I'd encourage you to start with a simple concept and see how our agents break it down. What kind of project are you considering? I can provide more specific technical details based on your use case.

Helena
MGX looks super interesting! How does it handle collaboration between different ai agents to ensure efficiency and avoid redundant tasks?
Mike

@hehe_z 
Thank you for this insightful question! MGX leverages MetaGPT's proven multi-agent collaboration framework, which has garnered 45k+ stars on GitHub.

The key is our structured SOP (Standard Operating Procedure) system. Each AI agent has clearly defined roles and responsibilities - like a PM writing PRDs, an Architect designing system architecture, and Engineers implementing code. They communicate through a standardized protocol, sharing context and outputs seamlessly.

To prevent redundancy, we've implemented a task dependency system where agents coordinate their work based on clear prerequisites and deliverables. For example, the Engineer won't start coding until the Architect has finalized the technical design, ensuring efficient workflow.

Would love to hear your thoughts on this approach!

Alex

@hehe_z @zongze_x Correction: MetaGPT now has 48k stars.


MGX handles collaboration between different AI agents using a centralized design with a Team Leader who distributes tasks. Each agent has access to a Jira-like tool that helps prevent duplicate work and track progress across the development lifecycle.

The Team Leader serves as the coordinator, assigning specific responsibilities to specialized agents (Product Manager, Architect, Engineer, etc.) based on their expertise. This ensures that each part of the development process is handled by the most appropriate agent while maintaining overall project coherence.

We're planning to experiment with a more decentralized approach in the coming months, allowing agents to communicate freely with each other. While this could significantly increase token usage, it might also lead to more organic collaboration patterns similar to real development teams.

This balance between structured coordination and specialized expertise is what allows MGX to maintain efficiency while leveraging the unique strengths of each agent in the system. Have you tried using multi-agent systems for development before?

Bob

@hehe_z Thank you for your interest! As a core developer, I'm excited to share that our multi-agent collaboration is built on solid academic research. Our approach is based on the open-source MetaGPT project and our research paper "MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework" (https://arxiv.org/abs/2308.00352).

The key innovation in our research is a hierarchical collaboration mechanism where agents with different roles (PM, Architect, Developer) work in parallel while maintaining task coherence through a shared memory and state management system. This significantly reduces redundant work and ensures efficient task distribution.

Yan Bingbing

As a developer, I recently tried MGX, and I especially like its multi-intelligence collaboration model, which allows product managers, architects, engineers and other roles to be clearly defined, and makes the development process efficient and smooth. The only thing I wish to improve is the template selection and editing features, and I look forward to future versions that provide more templates and more flexible editing options.

Mike

@onbing Wow, great to hear your feedback! Actually, your needs are already covered in our 'Editor' feature - you can edit not just code, but also product documentation and technical design. Just hit save when you're done!

For templates, we're continuously expanding our collection. Meanwhile, you can explore our AppWorld to see all user-shared examples. Here are some popular ones you can instantly remix:

Just remix any of these and replace with your content - it's that simple! Would love to see what you create with these templates! 🚀

Bob

@onbing Thank you for your in-depth experience with our product. As the architect in the MGX team, regarding the template selection issue you reported, I will continue to optimize and deliver the best selection results. And in AppWorld (inside https://mgx.dev/), you can choose one to remix rather than start from scratch to get a good generation result.

Alex

@onbing Thank you for your feedback! I'm glad you've enjoyed MGX's multi-agent collaboration model. It's great to hear that the clearly defined roles of product manager, architect, and engineer create an efficient development process.

I have good news regarding your suggestions - we're actually planning to revamp the homepage templates within days, and template selection will be updated in the next week or two. If you encounter any template selection errors, you can start a new chat to try again with minimal token cost when you're in the early conversation stages.

For editing functionality, you can currently use the remix feature as an alternative. You can select specific versions to remix either during the conversation or from the playback timeline's versions feature. We also have plans to introduce more convenient in-conversation editing options in future updates.

We appreciate your input as we continue to improve MGX. Is there anything specific about the template selection or editing functionality that you'd like to see in these upcoming updates?

David
Maker

@onbing Thank you for your thoughtful feedback! 🙌 We’re thrilled to hear that the multi-intelligence collaboration model is helping streamline your development process. Regarding template selection, we're actually planning to revamp the homepage templates in the next few days, and template selection will be updated in the next week or two. If you encounter any issues with template selection, you can always start a new chat to try again with minimal token cost during the early stages of the conversation.

As for editing functionality, we currently have the remix feature, which allows you to select specific versions to remix either during the conversation or via the playback timeline’s version feature. We’re also working on introducing more convenient in-conversation editing options in future updates. Your feedback is incredibly valuable as we continue to improve MGX—stay tuned for exciting updates! 🚀

David Martín Suárez

This sounds amazing! 🚀 We’ll definitely be trying it out with the Velada Genius team (we’re also launching today on PH!) — Congrats on the launch! 👏

Bob

@david_martin_suarez Wow, today is really a special day with so many great works being launched. Congratulations on the launch! 👏👏👏

David
Maker

@david_martin_suarez This sounds amazing! 🚀 Congrats on the launch! 🎉👏

Mike

@david_martin_suarez Thank you! And congratulations on your launch too! Wishing Velada Genius great success on Product Hunt today. Let's support each other in this journey! 🚀

Alan Rivera
Launching soon!

This is super interesting! Would it work in tandem with an actual developer where you can divide tasks between agents and the dev or is it more of a hands off where this is the full dev team?

Bob

@alan_rivera Thank you for your attention. You've raised a very important point about human-AI collaboration. Our starting point is to have the Agent Team automatically help you realize your dream with your chat, but we've also introduced a human-AI collaboration mode. When the agent encounters difficult decisions, it will seek human assistance and work better with humans to solve the problems encountered.

Mike

@alan_rivera 
Thank you for this great question! MGX is flexible and can work in both scenarios:

  1. As a collaborative tool alongside human developers - you can use specific agents for certain tasks (like having the PM agent create PRDs while developers focus on implementation) or use features like 'inspect' and 'editor' to enhance your workflow

  2. As a complete development team - where the AI agents handle the entire process from requirements to implementation

This flexibility allows teams to choose the integration level that best suits their needs. Would love to hear how you're planning to use it! 🚀

David
Maker

@alan_rivera Glad to see you're interested in MGX!

MGX works by taking user input, assigning tasks across multiple agents, and delivering the final output. But what makes it special? The entire process is open and transparent——you can watch the development unfold in real time!

Once the agents finish their work, you can easily review, edit, or inspect the output to get the best results. So whether you're looking for a one-stop solution or a productivity booster, MGX has you covered!

What’s your preference when it comes to AI programming tools? Would love to hear your thoughts!

Denis 🐝

This seems like a very helpful tool. I'm currently testing it out a bit, and I'm very impressed!

Denis 🐝

Update: I'm so incredibly impressed, holy! This is amazing—I just created a full blog with content and 7 posts using just a simple 10-word prompt. This is the best one I've tested so far.

Mike

@denisss 
Thank you so much for your enthusiasm and for taking the time to share your experience! We're thrilled to see you exploring MGX's capabilities.

The blog creation example you mentioned is exactly the kind of efficiency we aim to deliver. If you enjoyed that, you might also like to try:

Creating an interactive portfolio (like https://alex-portfolio-yhx5c3-v1.mgx.world)

Building a data dashboard (https://chart-js-dashboard-bbn89v-v1.mgx.world)

Or even a game (https://flappysnake-v2.mgx.world)

Would love to hear more about your blog project and what features you'd like to see next! 🚀

Alex

@denisss Thank you for sharing your feedback - it's truly an honor to have users like you using our system. While MGX still has areas for improvement, particularly around mastering the remix functionality and modular decomposition to fully utilize its capabilities, we believe we're on the right path.


We're committed to continually enhancing the text leverage that MGX provides, making it even more powerful for users. Your experience validates that we're moving in the right direction with our multi-agent approach.


Would you mind sharing the blog you created to App World? Your insights help us understand how people are using MGX in creative ways!

David
Maker

@denisss Thanks for your awesome feedback!

MGX is designed to help everyone build their own powerful software team—not just developers! Whether you're in tech, design, business, or any other industry, you can effortlessly create your own personal blog, portfolio, digital business card, or even data dashboards and custom tools to support your work.

Plus, MGX integrates Supabase for persistent data storage, so you never have to worry about losing your data or privacy issues.

And there’s so much more to explore—can't wait to see what you build!

Xenia

MGX really intrigues me! The notion of transforming AI agents into a full-fledged dev team is groundbreaking. I’d love to hear about any cool projects that have already been built using MGX!

Mike

@xenia1234 Thank you for your interest! We're excited to share some projects built with MGX. Our users have created everything from interactive games (like a Flappy Bird-inspired game at flappysnake-v2.mgx.world) to professional portfolios (alex-portfolio-yhx5c3-v1.mgx.world).


What makes MGX special is how our AI agents collaborate like a real dev team - from planning and design to testing and deployment. Each agent brings specialized expertise, just like human developers.

Would love to hear what kind of project you're interested in building! Feel free to check out our Discord community where users share their experiences and creations: discord.gg/ZRHeExS6xv

Alex

@xenia1234 I'm excited to see your interest in MGX!

For examples of cool projects built with MGX, I'd recommend checking out the App World section on our homepage. All the projects there are completely transparent - you can view the entire conversation history and prompts used to create each application.

App World showcases a variety of projects across different domains, giving you insight into what's possible with our multi-agent approach. You can explore how different teams of AI agents collaborated to build various web applications from start to finish.

The transparency allows you to see exactly how the Product Manager, Architect, Engineers, and other agents worked together to transform initial ideas into functional products. This can be incredibly helpful for understanding the capabilities and workflow of MGX before you start your own project.

Is there a particular type of application you're interested in building with MGX?

David
Maker

@xenia1234 MGX excels at building web apps across various scenarios! Here are some popular demos for different use cases:

Art Portfolio: https://art-portfolio-xn11wn-v3.mgx.world

Fun Interactive Game: https://flappysnake-v2.mgx.world

Data Visualization: https://chart-js-dashboard-bbn89v-v1.mgx.world

Personal Website: https://alex-portfolio-yhx5c3-v1.mgx.world

And who knows? There could be even more possibilities, depending on your ideas—because every idea is coconut-level cool! Can't wait to see what you create!

James Cooper

MGX is a powerful and flexible platform, well-suited for users who require multiple-role collaboration. Good luck to it!

Mike

@focusaur 
Thank you for your kind words! Yes, multi-role collaboration is exactly what we've focused on with MetaGPT's core capabilities. Your support means a lot to us as we continue improving the platform. 🙏

Alex

@focusaur Thank you for your encouragement!

MGX is indeed designed to leverage the power of multi-role collaboration, bringing together specialized agents to handle different aspects of development. We appreciate your support as we continue to develop and refine the platform.

If you have any questions or want to explore MGX further in the future, please don't hesitate to reach out. Wishing you all the best with your projects!

David
Maker

@focusaur Thanks for your interest!

Each MGX agent has unique strengths:

Mike – Task coordination & management

Emma – Product design & documentation

Bob – System architecture

David – Data processing & analysis

Alex – Coding & development

Together, they form your 24/7 AI-powered software team!

We’d love for you to explore more MGX features and share your valuable feedback. Can't wait to hear your thoughts!

Manish Choudhary

This is game changer. Would love to test it out, what is the best use to try this for? @zongze_x

Mike

@manish_choudhary19 Thank you for your excitement! For first-time users, we recommend starting with web development projects - they're a great way to experience our AI team's capabilities.

Check out some popular examples that others have built:

The best part? You can instantly remix any of these projects and customize them with your content. What kind of project interests you most? We'd love to see your creation! 🚀

Bob

@manish_choudhary19 Thank you for your love of our product. We have some video tutorials available at https://deepwisdom.notion.site/mgx-help-center, which we hope will help you get started faster and create your own unique applications.

Shivam Singh

MGX certainly seems to break new ground by incorporating multiple specialized agents to cover diverse aspects like data handling, coding, and design. It’s fascinating to think about how these agents can collaborate in real-time to streamline the creation process.


Congrats on the launch!
Best wishes and sending lots of wins :)

Mike

@whatshivamdo 
Thank you for your thoughtful observation! Indeed, MGX is powered by MetaGPT - our open-source multi-agent framework that's been battle-tested by the developer community with 45k+ stars on GitHub.

We've built MGX as a user-friendly platform that harnesses MetaGPT's core capabilities in multi-agent collaboration. Each specialized agent (PM, Architect, Engineer) follows structured SOPs to work together, just like a real development team.

Your support means a lot as we continue to evolve both MetaGPT and MGX. Feel free to join our Discord community (discord.gg/ZRHeExS6xv) to explore more! 🙏

David
Maker

@whatshivamdo Thanks for your enthusiasm! We’d love to invite you to explore and follow AppWorld, where you can check out all the amazing user-shared projects. It’s inspiring to see everyone’s creativity—because every user is a product designer!

Here are some popular template examples:

Art Portfolio: https://art-portfolio-xn11wn-v3.mgx.world

Link Hub: https://linkhub-7rkvvs-v1.mgx.world

Professional Portfolio: https://alex-portfolio-yhx5c3-v1.mgx.world

Not sure where to start with your first project? Just Remix any of these and swap in your own content! We can’t wait to see what awesome apps you create and share!

Bertram Ray

MGX is a bold experiment, attempting to replace costly human labor in software development with AI agents. I’ve used MGX for some time now, but the current version is far from mature—issues like excessive token consumption, a development process that struggles to converge (a problem exacerbated by MGX’s reliance solely on natural language interaction), and its apparent unpreparedness to handle a sizable user base have created friction. Server instability and a fragile payment mechanism drive many users to churn quickly after onboarding. While positioned against Devin, MGX falls short when compared to tools like Cursor and Trae, lacking the polish to compete. That said, for a startup team, MGX has already managed to pique user interest. Many of its current shortcomings will likely be addressed over time. It’s encouraging to see the MetaGPT team’s relentless efforts in this space, and I’m confident they’ll improve. The natural language programming arena is vast enough to accommodate multiple contenders.

Mike

@bertramray 
Thank you for this incredibly detailed and constructive feedback! Your insights are invaluable for our continuous improvement journey.

We've been actively addressing the points you raised through several iterations. Token consumption has been significantly optimized, server infrastructure expanded, and concurrent processing enhanced to provide a more stable experience.

Regarding the comparison with Cursor and Trae - you're right that MGX takes a distinct approach by focusing purely on natural language programming. This is a deliberate choice based on our conviction that natural language will be the future of human-AI collaboration in development. While this path presents unique challenges, it also opens up new possibilities.

We've also introduced practical features to bridge the gap between natural language and traditional coding:

  • The 'inspect' function enables one-click resource switching and immediate access to element code references for targeted modifications

  • Our 'editor module' supports both direct code editing and documentation updates, providing flexibility when needed

Your observation about the market being large enough for multiple approaches resonates deeply with us. We're committed to our vision while continuously refining our execution based on valuable feedback like yours.

We'd love to hear more about your experience and specific use cases. Feel free to reach out to our team directly - your insights help shape our development priorities. 🙏

David
Maker

@bertramray Thank you for your continued deep dive into MGX—your insights are incredibly valuable! We're doubling down on refining the development process to improve convergence and boost success rates. Think of it as giving our AI a bit more ‘wisdom’ so it doesn’t burn through tokens like an over-caffeinated coder. ☕💨

To make MGX smarter, we’re working on embedding more project-level development insights into the system—helping it not just execute, but also think and learn like an experienced engineer. At the same time, we're optimizing trajectory self-learning and iteration during the development process, so MGX can adapt and improve dynamically.

We’re also excited to integrate more of our existing academic research into the system to level up performance such as Self-Supervised Prompt OptimizationFACT: Examining the Effectiveness of Iterative Context Rewriting for Multi-fact Retrieval and AFlow: Automating Agentic Workflow Generation.
Appreciate your patience and feedback—it keeps us on our toes and pushing forward!

Mia

Built a TikTok pet filter game during my coffee break ☕
The AI engineer roasted my JS code & rewrote it. Savage but effective 😂
10/10 would get owned by robots again.

David
Maker

@mia618601587321 Thank you for trying it out! 🚀 A coffee break is exactly the kind of moment we envisioned—turning quick bursts of inspiration into reality! 💡 And hey, getting your JS code rewritten by AI might be brutal, but at least it’s an efficient way to level up! 😂 Feel free to explore more MGX features here: https://deepwisdom.notion.site/mgx-help-center

Mike

@mia618601587321 
Love your sense of humor! 😄 Yes, our Engineer agent can be quite direct with code reviews - but hey, that's what good engineers do!

Would love to see your TikTok filter game in our App World - it could inspire others to build during their coffee breaks too! Just share it to App World and let the community check out your creation.

Can't wait to see what you build next! 🚀

flytothemars163

@zongze_x Sooo cute! My kids must love it. The enntry point is so interesting. Congratulations!

Bob

@flytothemars163 Thank you for your appreciation. We hope some games like https://flappysnake-v2.mgx.world can bring joy to you and your family. We also look forward to further discussions with you in our Discord community (https://discord.gg/ZRHeExS6xv).

Mike

@flytothemars163 
Thank you for the lovely comment! We're delighted that the friendly AI team members caught your attention. Making advanced technology accessible and engaging is exactly what we aimed for with MGX - even students can understand and get excited about building things! Really appreciate your support! 😊

Alex

@flytothemars163 Thank you so much for your kind words! I'm thrilled to hear you find it cute and that the entry point caught your interest. It's especially wonderful to know it might appeal to kids too! We put a lot of thought into creating an engaging and approachable experience.

Would love to hear more about which aspects specifically resonated with you or your children if you end up sharing it with them. Thanks again for the support and congratulations!

Ayush Limbad

This looks interesting. Congrats on the launch👏

I would love to have a conversation with the makers on how we can scale this on X & instagram.

David
Maker

@aayush_limbad Thanks, Ayush! 🚀 We'd love to explore how we can scale this on X & Instagram!

Mike

@aayush_limbad 
Thank you for the congratulations and your interest in scaling! Really appreciate the support. Feel free to DM me on X (@MetaGPT_) or reach out through our Discord community - would love to discuss potential collaboration opportunities! 🚀

Ajay Sahoo

This AI-driven development will surely help in creation of SOPs collaboration seamless revolutionizes product building. All the best for launch.

Mike

@ajay27324 Thank you for recognizing our focus on collaborative development! You've captured exactly what we're aiming for with MetaGPT's SOP-driven multi-agent system. Your support means a lot to us! 🙏

David
Maker

@ajay27324 

Thanks for your interest! We’d love for you to dive deeper into MGX and explore its full potential.

Join our Discord community(https://discord.gg/ZRHeExS6xv) to connect, share feedback, and discuss your experiences!

Looking forward to your insights and creations!

Kay Kwak
Launching soon!

Wow, it's amazing that MGX is covering the entire development lifecycle! How customizable is the web user create? Is it easy to make changes?

Mike

@kay_arkain 
Thank you for your interest! Yes, MGX makes customization really straightforward. We've built powerful features to help you:

Using 'inspect', you can easily switch resources and get element code references for targeted modifications. The 'editor' module allows you to directly edit both code and documentation on the fly.

These tools, powered by MetaGPT's multi-agent system, make it simple to iterate and refine your projects. Feel free to try it out and let us know what you think! 🚀

David
Maker

@kay_arkain Thank you for the great feedback! 😊 We're super excited that you're impressed by MGX covering the entire development lifecycle!

When it comes to customizing the web user interface, it’s definitely flexible. You can tweak and adjust the design, content, and structure as you go along, especially with the remix feature. It allows you to select previous versions and make changes based on your evolving needs. Making adjustments is designed to be as easy as possible, but the more specific your inputs, the more tailored the results will be. We're also working on adding even more customization options in future updates to give you greater control over every aspect of the UI.

App World showcases a variety of projects across different domains, giving you insight into what’s possible with our multi-agent approach. You can explore how different teams of AI agents collaborated to build various web applications from start to finish.

Feel free to explore more MGX features here: MGX Help Center. We can’t wait to see what you create with MGX! 🚀