platform status · independent triage guide

Is Bolt down — or is it just your app?

The indicator below comes live from status.bolt.new. If Bolt reports an incident, the in-browser editor and AI builds are affected for everyone. Your deployed app is a different story: it typically runs on Netlify or another host with a Supabase or external backend, so a failed deploy, a crashing serverless function, or an exhausted API quota takes your app down while Bolt itself stays green.

Checking Bolt's official status… live from status.bolt.new

Nightlamp is not affiliated with Bolt and this is not the official status page. For platform incidents, the source of truth is status.bolt.new. This page exists for the other half of the question: what to do when the platform is green and your app still isn't working.

what their status page covers

What status.bolt.new does tell you

  • The bolt.new editor and AI code generation
  • In-browser WebContainer runtime incidents
  • Platform-side deployment/integration incidents
what it can't tell you

Bolt can be green while your app is down

Your last deploy failed and nobody noticed

A build error on Netlify (or wherever you publish) leaves the old version live — or worse, a broken one. Bolt's editor works fine; your production URL doesn't reflect what you think you shipped.

A serverless function is crashing in production

The function behind your contact form, checkout, or AI feature throws on an edge case the preview never hit. The static pages load, the core action fails — invisible to every status page.

An API key expired or hit its quota

AI apps lean hard on OpenAI/Anthropic/Stripe keys. When a key is rotated or the quota runs out, the app loads beautifully and the main feature returns errors.

Your database connections ran out

Supabase or Postgres connection limits, paused free-tier projects, or changed connection strings break reads and writes while both Bolt and your host report all-green.

five-minute triage

Platform problem or your problem? Run this checklist

  1. Check status.bolt.new — editor and build incidents are platform-wide.
  2. Check your hosting dashboard (Netlify, Vercel, etc.): did the last deploy actually succeed?
  3. Open your production URL in a private window and test the core action, not just the landing page.
  4. Check serverless/edge function logs for errors since the last deploy.
  5. Verify API keys and quotas for the services your app calls.
  6. Check your database dashboard for paused projects or connection-limit errors.

If sign-in emails are part of the broken flow, our free magic-link tester checks the full email round-trip in one click.

Checklist points at your own app? Our guide to fixing a Bolt app that broke after deploy walks through each failure symptom by symptom.

Stop checking status pages after the fact

Nightlamp watches your Bolt app's actual flows — checkout, signup, login emails, scheduled jobs — and a real engineer sends a plain-English diagnosis when one breaks, whether or not the platform is having an incident.

Start 14-day trial · no card
faq

Frequently asked questions

Bolt's status page is green but my app is down. Why?

status.bolt.new covers the editor and build platform, not your deployed app. Failed deploys on your host, crashing serverless functions, expired API keys, and database connection problems all break your app without appearing on Bolt's status page.

My Bolt app works in the editor but not at my live URL. What's wrong?

The editor preview and your deployed site are different environments. Missing environment variables on the host, a failed production build, or functions that only exist in preview are the usual causes. Compare your host's deploy logs against the working preview.

How do I get alerted when my Bolt app breaks but Bolt is up?

Monitor the deployed app's real flows — page loads, form submits, the API-backed feature your users actually pay for. Nightlamp probes those continuously and sends a plain-English diagnosis from a real engineer when something breaks.

Newsletter

One real no-code incident + the fix, monthly

A post-mortem from a real no-code outage — what broke, how it was found, and the fix — once a month. Confirming also gets you the no-code incident runbook pack.

Double opt-in. One-click unsubscribe. No spam, ever.