Build it all on the Neon platform.

A shared realtime chat: sign in, message everyone over WebSockets, attach images, and set a profile picture. A moderation agent screens every message, and tagging @neon summons an AI assistant that knows Neon (docs + agent skills) and replies right in the chat.

• Next.js frontend on Vercel (Neon Auth)
• Realtime WebSocket server on Neon Functions
• Moderation + @neon assistant agents via Neon AI Gateway
• Messages in Postgres, images & avatars in Object Storage

Paste any pricing URL and an AI agent reads it, untangles the tiers, and hands out two scores — pricing clarity and agent friendliness. Like Lighthouse, but for pricing pages.

• Next.js frontend on Vercel
• Judge agent on Neon Functions
• LLM calls via Neon AI Gateway
• Ratings cached in Neon Postgres

Upload photos of your teammates, @-mention them in a prompt, and an agent uses their faces as the starting point. Your whole team shares one collaborative gallery.

• Next.js frontend on Vercel
• Neon Auth — agent gated by JWT
• Image-gen agent on Neon Functions
• Assets in Object Storage, tables in Postgres

Fork any Neon branch to copy its Postgres database and object storage together, then explore tables, rows, and files. Mutate a fork, swap back to its parent — the parent is untouched. Click around; it’s really fast.

• TanStack Router SPA on Vercel
• JSON API on Neon Functions
• Image-gen agent on Neon Functions
• Assets in Object Storage, tables in Postgres