14 days to build. Voice input, AI extraction, relationship intelligence layer , the whole thing.

The idea: walk out of any important conversation, vent for 60 seconds, and let AI pull out everything that matters. Prices quoted. Commitments made. The personal detail they dropped in passing. All of it structured and surfaced before the next conversation.

Here's what surprised me building this:

Voice input is a completely different product than text input

Generating an MVP with AI agents feels like magic until you hit day 3, when components start drifting out of sync.

When you're prompting full-stack features, agents are great at adding code, but notoriously bad at cleaning up after themselves. A few patterns I keep running into:

1. Redundant state loops: An agent creates three different local state hooks to track the exact same data across siblings, causing weird hydration mismatches and flickering UI.

Last week I shipped a release and checked it twice. The registry had the new version. The pinned CDN link served the new file.
The link people actually install from was still serving the old one. No error anywhere.
A green check only tells you about the thing it checked.
What's the one thing you still check yourself, even when the agent and the tests both say it's fine?

So I've been messing around with cursor and v0 all week trying to spin up a quick prototype.

Honestly the first hour is pure magic. Prompt a bit and boom u got a working UI. But man... once a real bug hits or the context window gets messy it all goes downhill fast. Literally spent like 3 hours today trying to fix code I didn't even write myself. Absolute flow state killer.

How are you guys actually pushing past the messy prototype phase and into shipping?

Do you drop the AI tools entirely once the core layout is done, or do you just keep a super strict boilerplate to stop the LLM from hallucinating everywhere? Let me know because I keep getting stuck right at the finish line here.

Hey everyone,

I don't come from a formal engineering background, but I've been experimenting with how far pure natural language prompting can push a functional full-stack app.

I recently shipped an interactive daily bidding leaderboard where users displace ranks dynamically, resetting every midnight UTC.

Hey Maker community!

I ve been experimenting with the "vibe coding" workflow lately, and there s something incredibly satisfying about building useful micro-tools using nothing but a single `index.html` file in VS Code and a web browser. No npm install, no complex frameworks, no heavy configurations.

To prove the concept, I just built a quick "SVG to React Component Converter" completely client-side in vanilla JS. It handles the drag-and-drop, cleans Figma metadata, fixes camelCase attributes, and even displays a real-time preview of the icon.

It reminded me that sometimes we overengineer things when a simple script is all we need to solve a micro-pain.

Hey makers

I ve been building an AI productivity platform over the past few months, and one of the biggest challenges has been deciding what users actually need versus what simply sounds like a good feature.

I started with focused tools for everyday tasks like writing, business planning, marketing, SEO, and productivity. As the product grew, I realized that adding more tools doesn t automatically make the experience better.

Now I m focusing more on making the overall workflow simpler and understanding which features people actually return for.

At @UXPin we've just deployed the prompt enhancer for our AI component creator.
From now on, short prompts will be evaluated and refined if the AI considers them too weak. This aims to improve the AI output - with prototypes returned by the AI more detailed and diverse.

Example and output below:

Original prompt: 'news portal'

Refined prompt:

In my case, the first app I vibe-coded, I learned that it is important for you to research market to see whether demand exists or not. Any app without demand and no low to no budget while starting will not work.

The lessons from second app was that in an extremely competitive market, if you don't have a differentiator, the idea simply will not work. Especially if you are creating a saas which many devs have already made and distributed before also.