Nimbus Template: A Notes App You Can Rework in Chat

Nimbus Template: A Notes App You Can Rework in Chat

Nimbus is one of the 25 templates in Zugo's gallery: a notes app with instant capture, live search and pinning, wrapped in a landing page and a sign in screen. Cloning it into your own project is free. After that you change anything by describing it in chat, and each edit costs 3 credits.

Below is what the template actually contains, where the working parts stop and the decorative parts begin, and the shortest route from the demo copy to something that belongs to you.

What is actually inside the Nimbus template?

Nimbus is three connected surfaces rather than one page, and knowing that before your first edit saves you from asking for changes on the wrong screen. There is a public page that sells the idea, a sign in screen, and the app itself behind it.

The four feature blocks on the landing page are written as behaviors rather than features: capture before the thought escapes, search that narrows while you type, pinned notes that stay on top, and just enough structure to keep the page readable. That framing is worth keeping even when you swap the product, because a reader recognizes a described habit faster than a bullet list of capabilities.

The empty states deserve a mention. A fresh project opens with zero notes and the interface still looks intentional instead of broken. Beginners usually forget that state entirely, then pay for an extra edit to add it later.

Which parts really work and which are only screens?

The notes themselves are real. A note you create survives a page reload, search runs against titles and bodies, pinning moves an item to the top, and the word count updates as you type. None of that is a picture of an app.

The sign in screen is a different story. It shows what an account flow looks like, but it does not check a password or separate one person's notes from another's. Everything lives in the browser where you typed it, so a second device opens an empty library.

That boundary matters on day one, not later. A private notebook on your own laptop works immediately. Shared notes for a team need a real database and real accounts, which you add on top with the Supabase connector rather than with another prompt. The mechanics of that step are covered in can AI build a database app.

You can verify the boundary in two minutes. Add two notes, open the published project in a different browser, and confirm the library is empty there. Doing that early prevents a much more expensive misunderstanding later.

Who is Nimbus a good starting point for?

It fits any single object with a body of text, a date, and a flag for importance. Personal notes, a reading journal, a snippet library, call logs, an idea list, session notes for a coach or therapist: all of these drop into the existing structure with barely any restructuring.

It fits badly when your data has several linked objects. Tasks that belong to projects with owners, deals that belong to companies, stock that belongs to suppliers: those need a different screen layout, and a closer template saves you money. Orbit covers a deal pipeline with contacts, and a task tool is closer to how to build a todo app with AI than to a notes app.

There is a third case that people underrate. You like the shell and want completely different content inside it. The list plus search plus detail pattern survives a change of subject perfectly well, so recipes, workouts, or client records settle into it without a fight.

Ask yourself one question: does one of your records fit on a single card with a title, a body, and a date? If yes, Nimbus is close. If a record needs attachments, nested items, and links to other records, the shell starts working against you.

How do you turn Nimbus into your own app?

Order matters, because meaning is cheaper to change before appearance. Change what the app is, then what it looks like, and you avoid paying twice for the same screen.

1. Rename the product and rewrite the promise. One edit replaces the name everywhere and swaps the hero line for yours.
2. Change the subject. "Replace notes with recipes: each card gets cook time and servings."
3. Adjust the screens. Add a tag filter, a sort by date, an extra counter in the sidebar, or remove what you never use.
4. Then style it. Palette, typeface, spacing. Doing this last means you style the layout you kept, not the one you threw away.
5. Publish and open it on a phone. The project gets an address like yourname.zugo.run, and a custom domain sits on top of that on paid plans.

Write each request as one specific sentence. "Make it nicer" costs the same 3 credits as "increase the line height in the note list and remove the card borders", but only the second one can be judged as done or not done.

Which edits do people ask for first?

The requests repeat enough to be worth keeping as a checklist. Tags with a sidebar filter. Sorting by last edited, with the date shown in the list. A delete button with confirmation. Search that covers the title and the body at once. Export of a single note as plain text.

Interface language is the other common one, and it is best handled in a single sweep: "translate the whole interface, including empty states, button labels, and the browser tab title." Empty states are the piece people forget, which is how an otherwise translated app ends up with an English "Nothing selected" in the middle of it.

Mobile behavior belongs at the end of the work, not the start. A two column layout has to become something else on a narrow screen, so say what you want: the list first, and the note opening full screen. Check the result on a real phone rather than a narrow browser window, because the two do not behave identically. The wider habit of writing testable requests is covered in how to edit after generation.

What does a reworked Nimbus cost?

Cloning is free, so the meter only starts when you begin changing things. A realistic first pass is one edit for the name and copy, one for the subject swap, and two or three for the screens.

Plans are simple: Free gives you 5 credits, Pro costs $25 a month for 200 credits, and Business costs $99 a month for 800 credits. In practice 200 credits is roughly 16 platforms, 33 builds, or 66 edits in a month, so reworking a template sits comfortably inside a single Pro month.

Here is the part usually left unsaid. The 5 free credits are not enough for a build from scratch, which costs 6. They are enough for a template, because the clone is free and one edit fits inside the allowance. If you want to try the product without paying, starting from the gallery is the only route that actually works.

Every build is opened in a sandbox before it reaches you, and a version that fails to load is not delivered. That removes the blank white page after an edit. It does not remove the other risk: a build can load correctly and still not match what you meant.

Where are the honest limits?

This is one app for one person. Nimbus assumes a single writer and a single browser. Roles, permissions, and shared libraries are a separate piece of work built on a real database and real sign in, not a phrasing trick in the prompt.

Specific logic arrives through edits, not one prompt. Encryption, cross device sync, and collaborative editing are described step by step and checked one at a time. Budget for that instead of hoping a single long request lands all three.

Zugo does not replace a development team on a complex product. A personal tool or an internal utility is fully covered here. Something you intend to sell and grow for years eventually outgrows a chat window, which is why the source can be exported to GitHub at any point: the project is yours to take.

The first step is small. Open the gallery at zugo.dev, clone Nimbus, and spend your first edit replacing notes with the thing you actually track every day. Within a couple of minutes you will know whether this shell is yours or whether a neighboring template fits better.