Nambut

• 

  design mode

• 

  view mode

Inspiration

I had the idea for Nambut long before this hackathon. I wanted to build a personal card that gave users more freedom to design and express themselves. However, after seeing similar ideas like Linktree, I put the project aside because I wasn't sure if it would solve a real problem.

When I discovered WebMCP and saw how AI agents could directly interact with tools provided by a website, I became interested in the idea again. I decided to bring Nambut back and explore how AI could help users create and customize their cards.

The name Nambut is a play on the Thai word "นามบัตร" (Nam-bat), which means a business card or a card containing your personal information.

What it does

Nambut is a flexible personal card builder where users can create, customize, and share their own cards.

Users can freely arrange elements, customize colors, rotate elements, apply effects such as glass blur and shadows, and choose from templates.

With WebMCP, AI agents can also interact directly with the editor using tools such as add_element, move_element, and delete_element.

How we built it

Nambut is a web-based visual editor with WebMCP integrated into its core functionality.

We exposed the editor's main actions as structured tools so that AI agents can understand and interact with the application directly, while keeping the normal UI available for human users.

Challenges we ran into

I had never built an application like this before, and this was also my first time seriously working with AI agents.

I had to learn how AI agents work and how to expose the tools I built in a way that an AI could understand and use correctly.

Learning about WebMCP and this new way of connecting AI with web applications was challenging, but it also gave me a lot more motivation to learn and build.

Accomplishments that we're proud of

I'm proud that I was able to bring an idea I had been thinking about for a long time back to life.

More importantly, I managed to finish it within the hackathon deadline and turn it into something that actually works.

What we learned

This project gave me my first real experience with AI agents and taught me that websites can provide AI with structured capabilities instead of simply being interfaces that AI has to navigate.

It also showed me how new technologies like WebMCP can change the way we think about building web applications.

What's next for Nambut

If Nambut proves that there is real potential, I want to add more tools and significantly improve the UX/UI.

I also want to make Nambut a way for more people to discover WebMCP and experience how AI can be used more effectively when it can directly interact with the tools of an application.

Built With

• express.js
• neondb
• node.js
• postgresql
• render
• vercel
• vite