How do I monitor a Lovable app after launch?

By Ohad Krispin, founder of VibeRaven · LinkedIn

Lovable gives every published app visit Analytics, and the built-in backend has Logs under More → Settings → Logs for Edge Functions, sign-ins, Postgres, Realtime and Storage, going back up to 5 days in the quick ranges. Project monitoring (beta, Pro and up) checks recent visitor errors and your code on a schedule and emails the project owner about important findings, but Lovable says it does not replace testing. For an alert with a stack trace when a browser error hits a user, add Sentry with @sentry/react; if your app uses your own Supabase project, its dashboard adds Logs, Reports and Advisors. In launch week, check errors, failed sign-ins, Edge Function errors, credit or database usage, and Stripe webhook deliveries once a day.

Repository check

Check this in your repo

Check the repo for error monitoring

For AI-built Vercel + Supabase apps, such as a Lovable project synced to GitHub, run the command below from the project folder with Node.js 20 or newer. When it finds no Sentry, PostHog or similar instrumentation in the code, it reports missing_monitoring, as info rather than a blocker; a Vite + Supabase fixture repo with no monitoring got that finding on 2026-10-29.

It is a repository check, not a live database test: it reads files, so it cannot see Lovable Analytics, Lovable Logs or project monitoring, and a run without the finding does not mean anyone is alerted when the app fails. It is advice, not a gate.

npx -y viberaven@1.6.12 check

Does Lovable have built-in analytics?

Yes. More → Analytics shows visits, page views, visit duration, bounce rate, traffic channels, pages and devices for a published app, on all plans, with no setup, and it sets no cookies. A visit is one browsing session, not one person, and custom ranges reach back 90 days. It counts traffic, not what users do inside the app or whether an action failed: for that, use product events (the PostHog connector) and error monitoring.

Where are the logs for a Lovable app?

On the built-in backend (Lovable Cloud), open More → Settings → Logs and pick a log type: Edge logs (the default), Function logs, Function edge logs, Authentication logs, PostgreSQL logs, Realtime logs and Storage logs; Server logs appear only on TanStack Start projects. You can search text, filter to error responses, and export what you loaded as JSON. Logs do not stream: select Refresh for new entries.

If the app is connected to your own Supabase project, the same events are in the Supabase dashboard under Logs, and per-function Invocations and Logs under Edge Functions. For which backend you are on, see Lovable Cloud vs your own Supabase.

Which Sentry guide fits a Lovable project?

Lovable projects built before May 13, 2026 are usually React + Vite; newer projects use TanStack Start, which renders pages on the server. Ask Lovable "What stack is this project on?" first. For React + Vite, follow Sentry's React guide (the steps below). For TanStack Start, Sentry has a separate TanStack Start React SDK, which Sentry marks as beta.

Checklist

Set up monitoring, then check it daily in launch week

01
Check the repo for error monitoring. For AI-built Vercel + Supabase apps, such as a Lovable project synced to GitHub, run the command below from the project folder with Node.js 20 or newer. When it finds no Sentry, PostHog or similar instrumentation in the code, it reports missing_monitoring, as info rather than a blocker; a Vite + Supabase fixture repo with no monitoring got that finding on 2026-10-29. It is a repository check, not a live database test: it reads files, so it cannot see Lovable Analytics, Lovable Logs or project monitoring, and a run without the finding does not mean anyone is alerted when the app fails.
```
npx -y viberaven@1.6.12 check
```
02Turn on Project monitoring (Pro, Business or Enterprise): Project settings → General → Project monitoring. The default runs daily at 6:00 only if the project was edited since the last check, so after launch set Runs to Every time. Each completed run uses credits; the owner gets email for important findings.
03
Add Sentry error monitoring to a React + Vite Lovable app. Error monitoring starts once Sentry.init() runs, and the init file must be the first import in your entry point. The DSN only allows sending events, so it can sit in browser code.
```
npm install @sentry/react --save

// src/instrument.ts
import * as Sentry from "@sentry/react";

Sentry.init({
  dsn: "https://<key>@o<orgId>.ingest.sentry.io/<projectId>",
});

// src/main.tsx: first line
import "./instrument";
```
04
Upload source maps so stack traces point at your code, not minified bundles. Sentry's wizard sets up the Vite plugin and build config; it needs a SENTRY_AUTH_TOKEN at build time.
```
npx @sentry/wizard@latest -i sourcemaps
```
05
Prove Sentry receives errors from the published app: add a button that throws, click it on the live URL (errors thrown from the browser console do not reach Sentry), and check the Issues page.
```
<button onClick={() => { throw new Error("Sentry Test Error"); }}>Test Sentry Error</button>
```
06Daily: open Sentry Issues for new errors, and More → Settings → Logs filtered to errors. Look at Function logs and Edge logs for failing Edge Functions, and Authentication logs for failed sign-ins and sign-ups. On your own Supabase project, the Auth and Edge Functions reports show auth errors and function status codes.
07Daily: open More → Usage to see what the project spends on Database, AI, Compute, Network, Storage and Realtime. Under Workspace settings → Usage limits & alerts, set a Run limit; a blocking limit pauses the built-in backend, published app included, from 80% of the threshold, so use notify only if the app must stay up.
08If the app uses your own Supabase project: check the organization usage page, and open Security Advisor and Performance Advisor after each migration. With the default email service, Supabase Auth sends 2 emails an hour and only to your team's addresses, so set up custom SMTP before real users sign up.
09If you take payments with webhooks: in Stripe Workbench → Webhooks, select your endpoint and open Event deliveries. Failed deliveries show the HTTP status; live mode retries for up to three days, and Resend works for 15 days. Lovable's Stripe integration uses no webhooks by default, so check whether your app added one.
10Add product analytics if you need to know what users do, not only that they visited: connect PostHog under Connectors with its phc_ project token (never a phx_ personal key), then ask Lovable to capture the events that matter, such as sign-up completed or checkout started.

Sources

Official docs behind this checklist

Browse all launch guides · Read resources · How to make and check your next change

FAQ

FAQ for builders before launch

Will Lovable tell me when my published app breaks?

Only if you turn on Project monitoring, which is in beta on Pro, Business and Enterprise. It runs on a schedule, not in real time, emails the project owner about important findings, and can miss issues or flag working behavior. For an alert when an error happens, add Sentry.

Do I need Sentry if Lovable already has logs?

The Logs page covers the backend: functions, auth, database, realtime and storage. A JavaScript error in a visitor's browser does not reach those logs. Sentry captures browser errors and uncaught exceptions with a stack trace, the user's page and an alert.

Can a usage limit take my app offline?

Yes, if you choose blocking. A blocking Cloud limit pauses the database, storage and backend services for the projects it covers, published apps included, from 80% of the threshold until the period resets or you change the limit. Adding credits does not lift it. A notify-only limit sends an alert and blocks nothing.

Where do failed sign-ins show up?

On Lovable Cloud, in More → Settings → Logs under Authentication logs. On your own Supabase project, in the Logs page with the Auth source, and in the Auth report, which charts sign-ins, sign-ups and auth errors.