Lovable App Builder Prompts: 30 Full Apps and the Best Practices
Prompt builder · four steps · no account
Step 1 of 4
Which one is closest?
Building something else? Describe it to the AI generator
A one-line prompt makes Lovable produce every screen of a support tool or a booking system, and none of them is wired to data that survives a refresh. These 30 prompts are whole small apps with the build order inside them. None has been run yet; this page says what the first run will record and what the three runs elsewhere in the library already show.
By Marco Kohns · Prompts written and run by us
Short answers
Short answers
Where do I start?
Match the data shape, never the industry label. A habit tracker, a reading list and a notes tool are the same shape; pick the closest prompt and change the nouns.
Pick by shape →What does a run cost?
? credits
Not measured for this category yet. The three runs we have measured elsewhere cost 1.8, 9.5 and 10.4 credits; a whole-app prompt with a backend sits at the top of that range.
See what we do know →Build mode or Plan mode?
Build mode, with the prompt's closing line making Lovable ask first. Plan mode is for after: scoping a larger feature or a bug that survived 2 fixes. It costs 1 credit per message and never edits code.
Which mode when →My app exists and looks finished
Then check one read path against live data. The common failure is screens rendering placeholders; the fix lines below point an existing app at a table.
Fix lines for live apps →30
App Builder prompts in this category
↳ Copy-paste, free
0of 30
Prompts from this category run in Lovable
↳ Next run: Customer Support Ticketing Tool
3runs
Measured so far across the library
↳ 1.8 to 10.4 credits each
35
Upvotes since Jan 2026
↳ One per visitor, self-reported
Data basis: no run in this category yet; every card says so. The credit figures on this page come from the three measured runs elsewhere in the library, read from the Lovable balance before and after. Mode and credit rules from Lovable's docs, checked 22 October 2026.
Find a prompt
Which shape of app are you building?
Cards are sorted by build order. Each says whether we ran it in Lovable, and if not, shows a question mark instead of a guess.
Showing 30 prompts
- Step 1 · One read path▲ 8
Internal Prompt Library Browser
An internal prompt library where each prompt declares its variables, Copy fills them through a form, and row-level security keeps every workspace's prompts invisible to outsiders.
Run it when: you want the smallest app that renders live data, to learn the loop on.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 1 · One read path
Simple Blog CMS
A blog CMS with markdown, drafts and scheduling where slugs lock after first publish, and drafts return 404 and stay out of the feed and sitemap.
Run it when: the content is yours and the readers are not signed in.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 1 · One read path
Knowledge Base / Help Center
A public help center where every article has an owner and review date, search logs zero-result queries, and a health view lists stale, unhelpful and orphaned articles.
Run it when: the same questions arrive by email every week.
Not run yet · credits ? - Step 1 · One read path
Member Directory with Search
A members-only directory where each profile field has a visibility toggle, search reads only member-visible fields, and contact details stay hidden until a request is accepted.
Run it when: a list of people has to be searchable by more than name.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 1 · One read path
Employee Directory App
An employee directory with sign-in limited to the company email domain, search by expertise tags, and a now-working-on line that expires each quarter and asks for a refresh.
Run it when: the directory is internal and the org chart matters.
Not run yet · credits ? - Step 1 · One read path
Reading List Tracker
A private reading tracker where pasting a URL fetches title and site, abandoned is a status, finishing asks for a takeaway, and a weekly pick resurfaces 3 stale items.
Run it when: you want a personal tool and a first project in one.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 2 · Write path▲ 1
Contact Form Backend Storage
A contact form backend that stores every valid submission before notifying anyone, so a failed email marks the lead Notification failed and it still appears in the inbox.
Run it when: a form has to store what it receives, with the failure state specified.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 2 · Write path▲ 1
Form Builder with Conditional Logic
A form builder with one-level show-when rules, a live Debug preview, and submissions stored with the form version and the fields visible at submit time.
Run it when: the questions depend on earlier answers.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 2 · Write path
Survey Builder + Results Display
A survey tool with five question types, one question per screen on a phone, resumable answers, and per-question drop-off counts that show where respondents quit.
Run it when: you need the results screen as much as the form.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 2 · Write path
Simple Quiz Builder
A quiz builder for scored and outcome quizzes where publishing is blocked until score bands cover the range without gaps, and answer weights stay server-side until submission.
Run it when: answers are scored and the score is the product.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 2 · Write path▲ 1
Poll & Results Display
A one-question poll voters answer in one tap on a phone without an account, with live results rounded to sum to 100 and the duplicate-protection mode printed.
Run it when: one question, many voters, a live chart.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 2 · Write path▲ 1
File Upload + Preview
A reusable upload component with instant local previews, per-file progress, cancel and retry, and type and size checks that run in the browser and again on the server.
Run it when: users bring files and need to see them before saving.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 1
Kanban Board for Workflow Tracking
A kanban board where moves run through one backend function that checks WIP limits and entry rules, and every transition is recorded for cycle time and aging.
Run it when: anything has a status, an owner or an order.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 6
Customer Support Ticketing Tool
A ticketing tool for 2 to 15 agents where web forms and forwarded email become owned tickets, and internal notes sit in a table requesters can never read.
Run it when: requests arrive from outside and someone has to own each one.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 2
Habit Tracker App
A habit tracker with one-tap check-offs and streaks computed from each user's schedule and stored timezone, so weekend gaps on weekday habits never break a streak.
Run it when: streaks and timezones have to be right or the app is wrong.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 1
Task List with AI Suggestions
A task list where AI proposes subtasks, priority and time estimates per task, and nothing changes until the user presses Accept on each proposal.
Run it when: a plain task list is done and you want suggestions on top.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 1
Budget Planner Tool
A budget planner that stores money as integer cents, shows what is left per category each month, and lets each category reset or carry surplus and deficit.
Run it when: categories, periods and totals have to agree.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 2
Booking System (Date + Confirm)
A booking system for a solo service business where a unique constraint on service and start time makes double-booking impossible, and manage links honour the notice rules.
Run it when: a slot can be taken once and confirmation matters.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Step 3 · State model▲ 1
Appointment Availability Scheduler
An appointment scheduler that resolves slots from weekly rules and exceptions at request time, books in a transaction that re-checks the slot, and shows invitees their own timezone.
Run it when: availability is the product and double-booking is the bug.
Not run yet · credits ? - Step 3 · State model▲ 1
Resource Booking Manager
A booking manager for rooms, equipment and desks where a database exclusion constraint on resource and time range blocks conflicts, with per-resource rules and weekly utilization.
Run it when: rooms, desks or equipment get booked by a team.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Standalone▲ 4
Multi-User Chat Interface
Multi-user chat with channels and DMs where a reconnecting member receives every missed message once, in server order, and non-members cannot load history by direct request.
Run it when: delivery and history rules matter more than the UI.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Standalone▲ 1
AI Customer FAQ
An AI FAQ that answers only from your own entries with each one cited, admits a miss below a confidence threshold, and logs misses ranked by count.
Run it when: the answers exist in a document and the model should stay inside it.
Not run yet · credits ? - Standalone▲ 1
AI Email Template Builder
An AI email template builder that fills a fixed table-based skeleton from a one-line brief, keeps the footer locked, and blocks export when a merge field is unknown.
Run it when: the output is text someone will send.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Standalone▲ 1
AI Meeting Notes to To-Dos
Turns pasted meeting notes into proposed action items that quote their source line, name an owner only when the line does, and become tasks after one review.
Run it when: notes come in, to-dos with owners should come out.
Not run yet · credits ? - Standalone
AI Job Description Creator
A job description creator that generates six fixed sections from a short brief, caps must-haves at 6, and flags exclusionary phrases with the rule each flag cites.
Run it when: the output has a fixed structure the model must respect.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Standalone
AI Recipe Idea Generator
A recipe generator that returns 2 to 3 structured recipes from the ingredients you have, marks each item Have or Need, and discards any recipe containing a declared allergen.
Run it when: you want a small AI wrapper to learn the call shape on.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock - Standalone▲ 1
AI Event Recommendations
An event feed that ranks upcoming events from stored preferences and behaviour, shows a because-line on every card, and drops any event that has no reason to show.
Run it when: there is a catalogue and a history to recommend from.
Not run yet · credits ? - Step 4 · Aggregate views
Internal Tool Dashboard Builder
An internal dashboard builder where teams compose widgets from code-reviewed named queries with bounded parameters, and the browser sends only a source name and parameters.
Run it when: every other model is settled and someone decides from the numbers.
Not run yet · credits ? - Step 4 · Aggregate views
CRM Lite with Contact Management
A lightweight CRM for 2 to 10 reps where email is the duplicate key with a Merge flow, and Today lists overdue follow-ups first every morning.
Run it when: you want the one-prompt CRM before the ten briefs on the CRM page.
Not run yet · credits ?Pro version adds the growth-channel instructionsUnlock
Building more than one app? The bundle has all 30 as one kit in build order, with every Pro version unlocked, plus 122 more prompts.
See what is in the 152-prompt bundleDisclosure: if you create a Lovable account through links on this page, we may earn a commission. It costs you nothing and does not change what we recommend.
Measured run
No app prompt has been run yet. This is what the first run will record.
▍Context
Build a support ticketing tool for 2 to 15 agents: requests arrive by web form and forwarded email, become tickets with a status and an owner, and agents reply with internal notes alongside. The one job: no customer request gets lost between agents.
▍Core Features
1. Intake: a support form and an inbound email address both create tickets with the requester's identity; a reply to a resolved ticket reopens it
2. Lifecycle: new, open, waiting on customer, resolved; status changes automatically with reply direction, and waiting on customer pauses the response clock
3. Assignment: one owner per ticket, an unassigned queue visible to all, one-click Claim
4. Replies and notes: replies email the requester; internal notes render in a separate yellow block and can never be sent
5. Metrics: first-response and resolution time per agent and overall, with a soft target per priority highlighted on breach
▍Ticket Lifecycle & SLAs
▍Before Building
Ask me any questions you need in order to fully understand what I want from this feature and how I envision it. Save my answers to the project knowledge so later prompts build on them.
▍Distribution hooks · Pro
Locked. The Pro version adds the growth-channel instructions for this screen.
- 01
A fresh project in Marco's own workspace
Balance read and screenshotted first. Never a client workspace, never a reused project: a follow-up prompt in an old project is not comparable.
- 02
The prompt pasted unedited, in Build mode
The support ticketing prompt, the second most upvoted in this category. Its Before Building line should make Lovable ask about the team and the intake channels first.
- 03
Every question recorded, with the answer given
On the two runs that had this closing line, Lovable asked 2 and 4 questions before writing code. The count goes on the card.
- 04
Rounds counted until the Core Features list is met, or 5
Each follow-up prompt verbatim and why it was needed: a missing state, a broken path, a visual fix. A run that stops at 5 rounds is published as such.
- 05
The preview screenshotted at 1280 px and 390 px
Plus the Security view after a scan, which is free, and the code view to see what the backend got.
- 06
Balance read again
Credits used equals before minus after, to one decimal, with the day's 5 daily credits counted because they are spent first. Then the row goes into the runs file this page reads.




Credit math
What a whole-app run is likely to cost, from the runs we have
No figure for this category yet, so this table uses the two measured runs that provisioned a backend: 9.5 credits for the states system and 10.4 for the SaaS foundation. Lovable bills Build mode on the work done and says many requests cost less than one credit; a whole app is not one of them.
| Plan | Daily build credits | A backend-provisioning run (9.5 to 10.4) | Unused credits |
|---|---|---|---|
| Free | 5 a day, max 30 a month | Two days of daily credits plus part of a third. Two runs this size use most of the month's 30. | Daily credits do not roll over |
| Pro | 5 a day, plus 100 monthly plan credits from $25 | 5 from the day's credits, 4.5 to 5.4 from plan credits, on a fresh day | Plan credits roll over, expire after 2 months |
| Business | 5 a day, plus 100 monthly plan credits from $50 | Same split as Pro | Same as Pro |
Lovable's own examples, for scale
- “Make the button gray”
- 0.50
- “Remove the footer”
- 0.90
- “Add authentication”
- 1.20
- A Plan mode message
- 1.00 plus research
Source: Lovable docs, Credits and usage (Build examples marked illustrative) and Plan mode. Checked 22 October 2026.
Three ways to spend less
- Refine in Chat before you build. Lovable detects a detailed prompt in the dashboard box and offers a free chat to refine it. Chat mode cannot edit files and costs a fraction of a credit per message after the daily allowance.
- Fill the knowledge file once. Purpose, personas, schema and constraints, up to 10,000 characters, included in every message. The Context and rules sections of the prompt you pick are the first draft.
- Switch to Plan mode after 2 misses. A plan message costs 1 credit plus research and changes nothing until you approve it. A fourth Build attempt at the same bug costs more and fixes less.
Start here
Which app prompt should you run first?
Pick the sentence that sounds like your project.
One read path, then one write path
- 01Internal Prompt Library Browser · a list rendering live data
- 02Contact Form Backend Storage · a form that stores what it gets
- 03Kanban Board for Workflow Tracking · the first state model
Build order
Four steps. One piece at a time, checked before the next.
01
One read path
A list on live data
The smallest screen that renders rows from a table. Everything else is a variation of it.
02
Write path
A form that stores
Validation and the failure state specified with the form, never after it.
03
State model
Status · Owner · Order
Anything with a status needs this before more screens. Kanban, tickets, bookings.
04
Aggregate views
Dashboards last
They depend on every other model being settled.
Anatomy
What each section of an app prompt does
| Section | What it tells Lovable | What you get without it |
|---|---|---|
| Context | Who uses the app and what they do in it, in one line | The first question Lovable asks, or a guess |
| Core features | The pieces, numbered, in the order they should exist | Every screen at once, none wired to data |
| Rules section | The lifecycle, the slot rule, the WIP limit: what a screen cannot show | A status column nothing enforces |
| Build order | One read path end to end, then the write path, then the rest | An interface Lovable cannot then wire up |
| Safe-guards | What stays untouched and what never leaves the server | Regressions in the screen that worked |
| Before building | Ask questions first, save answers to project knowledge | The 2 to 4 questions arrive as guesses instead |
A one-line prompt gets you
- Every screen, none of them finished
- A data model optimised for one working screen
- Status fields nothing enforces
- A fourth Build attempt at the same bug
A prompt from this page gets you
- A build order Lovable can follow in one message
- Rules a screen cannot show, written down
- Questions first, then the build
- A schema you read before the second feature
Fixes
Where app prompts go wrong, and the line that fixes it
Already built an app that looks finished? Paste the fix line into your existing project.
Symptom
Every screen renders placeholder data
Cause: the whole app was asked for in one sentence, so Lovable produced the screens and left the data layer for later.
Fix line
Take the [main list] screen only. Create its table, seed 3 rows, render them from the database, and show me the empty state when the table is cleared. Do not touch any other screen.
Symptom
The status changes and nothing happens
Cause: a status column was requested with no rule about what a change does.
Fix line
When status changes to [resolved], set resolved_at, stop the response clock and write a history row with who changed it. Reject any transition not in this list: [states].
Symptom
The same fix fails for the third time
Cause: Build mode is being asked to guess at a cause it has not investigated.
Fix line
Switch to Plan mode. Investigate why [symptom] happens, list the files involved and the probable cause, and propose the smallest change. Do not edit anything until I approve the plan.
Symptom
The second feature broke the first
Cause: the second prompt had no safe-guard line, so the model refactored what already worked.
Fix line
Add [feature] without changing the existing [screen] or its table. If a change there is unavoidable, stop and tell me what and why before making it.
Upvotes
Which App Builder prompts do builders upvote?
Upvotes since January 2026, one per visitor. Small numbers: read them as a hint, never as a ranking.
- Internal Prompt Library Browser8
- Customer Support Ticketing Tool6
- Multi-User Chat Interface4
- Booking System (Date + Confirm)2
- Habit Tracker App2
- AI Event Recommendations1
The Complete Lovable Bundle
All 30 app prompts in build order, plus every other prompt
Every prompt and kit, every Pro version unlocked. One payment, no subscription.
One payment. No subscription, no renewal, ever.
FAQ
Questions about App Builder prompts
What are the best practices for prompting the Lovable app builder?+×
Lovable's own prompting guide, read on 22 October 2026, is organised in four phases: lay the foundation, think in systems, build with precision, iterate and ship. Its rule for a change prompt is what to build, where it goes and what must stay untouched. The prompts here encode that as a Context line, numbered features, a rules section, a build order, safe-guards and a closing line that makes Lovable ask before building.
Should I prompt the whole app at once?+×
The docs say no: a full-page prompt gets you noise, a section-based prompt gets you signal. Each prompt here is one message, but its Build Order section tells Lovable which piece comes first, which is what makes one paste behave like a sequence. Two prompts from this category run one after the other on one project, never concatenated.
Build mode, Chat mode or Plan mode?+×
Build mode for the prompts here. Chat mode cannot edit files and costs a fraction of a credit after a daily allowance, so it is for refining a prompt first. Plan mode costs 1 credit per message plus research, never modifies code, and is for scoping a larger feature or a bug that survived two fixes. All three from Lovable's docs, 22 October 2026.
Which of these prompts have been run?+×
None of the 30 yet, and the cards say so. The three measured runs in the library are a landing page (1.8 credits, one round), a UI states system (9.5, 2 questions first) and a SaaS foundation (10.4, 4 questions first), all in fresh projects in our own workspace. The first app-builder run gets the same ledger.
What goes in the knowledge file for an app like these?+×
Lovable's docs recommend a one-page brief, up to 10,000 characters, included in every message: the app's purpose, personas, schema, constraints and design rules. The Context and rules sections of the prompt you pick are the first draft of that file, and the answers Lovable collects before building are saved there too.
Are these Lovable app templates?+×
They are prompts that generate the app inside your own project. For this category that matters: a template fixes a data model up front, and thirty app shapes need thirty different models. Each prompt's rules section implies the model for its app, and you read it before adding the second feature, while changing it is still cheap.
How long should a Lovable prompt be?+×
Long enough to carry priority, build order and constraints, which is a few hundred words here. The length is in the sections, never in adjectives: Lovable's docs say to describe components and behaviour, and a word like modern is invisible to the model as an instruction.
What we have not tested yet
- All 30 prompts in this category. Their cards show a question mark until we run them.
- The same app prompt in Build mode and through Plan mode first, with the credit difference.
- Two prompts from this category run in sequence on one project.
- Runs on Bolt, v0 or Base44. Only Lovable runs are ours.
Next run: Customer Support Ticketing Tool, the prompt excerpted above. How we test prompts
Who wrote this
Marco Kohns wrote these prompts and ran the three measured ones elsewhere in the library. Lovable Prompts is independent and not affiliated with Lovable.
“Try in Lovable” links are affiliate links: we may earn a commission when you create an account through them. The Pro versions and the bundle are our own paid products. Neither decides which prompt we tell you to start with.
Related reading
- the Lovable app builder guide · A tour of the editor with our screenshots.
- UI prompts for the shell and states · Navbar, dialogs, tables, empties.
- backend prompts · Roles, jobs and recovery for a data-backed app.
- the CRM build kit · Contacts, deals and RLS that blocks reassignment.
- Lovable pricing and credits · Plans read today and what one prompt costs in credits.
- Lovable alternatives compared · Six tools, prices re-verified.
- the Lovable prompting guide · Channel first, then the prompt.
- getting the first 100 users · What to do once the app runs.
- from idea to app, measured · Brief, prompt, build, publish, first users.
- Base44 alternatives · If the app is moving from another builder.
- Lovable mobile app · What a Lovable app can be on a phone, from the docs.
- app ideas · The thirty apps as ideas, grouped by channel and ranked by upvotes.
- SaaS ideas · Fifteen of these apps as SaaS, with the mechanics that make them chargeable.
- CRM app prompts · CRM Lite expanded into ten briefs on a proven schema.
- what is an AI app builder · How a prompt becomes an app with a backend, measured.
- Lovable prompt generator · Writes a structured prompt for an idea that matches no library entry.