Open Source · MIT Licensed

Your AI writes
your code.
ScanWarp keeps
it running.

One command. ScanWarp scans your app, aggregates alerts from your entire stack, and when something breaks, AI diagnoses the problem in plain English. Your coding tool already knows the fix.

$ npx scanwarp init
? What's your app URL? https://my-app.vercel.app
✓ Scanning... found 4 pages, 2 API routes
✓ Vercel log drain connected
✓ Stripe webhook listener active
✓ Supabase health monitoring on
✓ MCP configured for Claude Code
✓ ScanWarp is watching.
Join Waitlist View Source
▼ Scroll
The Problem

You built your app by prompting. You find out it's broken from a DM.

You described what you wanted. Cursor built it. You deployed to Vercel. Users showed up. Revenue started flowing. Then this happened:

MON 2:00PM Deployed discount code feature
MON 2:14PM User enters invalid code
MON 2:14PM Checkout page crashes — unhandled exception
MON–FRI 47 users hit the same bug. 0 complete checkout.
FRI 9:22AM You find out from a Twitter DM: "your site is broken"

You looked at Datadog. You saw dashboards, P99 latencies, log aggregation, APM configuration. You closed the tab. You built your entire app in English. Why can't monitoring work the same way?

With ScanWarp

Same scenario. Different outcome.

MON 2:14 PM
ScanWarp detects crash
Vercel log drain catches the 500 error on /api/checkout. Stripe reports payment_intent.failed. ScanWarp correlates both — same root cause.
MON 2:15 PM
AI diagnoses the problem
"Your discount code handler throws an unhandled exception when the code doesn't exist in the database. The checkout page crashes for any invalid code."
MON 2:15 PM
You get a Discord alert
Plain English. What's broken, why, and a fix prompt ready to paste.
MON 2:17 PM
You open your editor
Your AI already knows. It read the diagnosis through MCP. You say "fix it."
MON 2:19 PM
Deployed. Confirmed working.
ScanWarp runs the check again. Green. Total downtime: 5 minutes.
SW
ScanWarp 2:15 PM
⬥ CRITICAL — checkout is down

What: /api/checkout returning 500
Why: Discount code query fails on invalid codes — no null check on the Supabase response
Impact: 3 failed payments in the last 60 seconds
Correlated: Stripe payment_intent.failed ×3

Fix: Add null check in /api/apply-discount before accessing row.discount_percent
What It Scans

Every provider. Every signal. One feed.

ScanWarp connects to your entire stack and aggregates everything into one unified stream. No more checking five dashboards.

External Monitoring

Hits every page and API route every 60 seconds. Catches broken pages, slow loads, JS errors — everything a user would notice.

Provider Alerts

Ingests logs and webhooks from Vercel, Stripe, Supabase, Cloudflare, and more. Catches issues before they surface externally.

Correlation Engine

A Stripe failure and a Vercel 500 in the same minute? ScanWarp groups them as one incident with one root cause.

AI Diagnosis

Every incident gets a plain English explanation. What broke, why, and exactly how to fix it. No stack traces to read.

MCP Integration

Your AI coding tool reads ScanWarp through MCP. When you open Cursor or Claude Code, the AI already knows what's wrong.

Is It Me Or Them?

Checks provider status pages automatically. When Vercel is having an outage, ScanWarp tells you it's not your code.

Connects To
Vercel
Supabase
Stripe
Cloudflare
GitHub
Resend
Railway
AWS
Honest Scope

What ScanWarp is not.

We'd rather you trust us than wonder later.

Not Analytics
That's PostHog

ScanWarp tells you your app works. PostHog tells you what people do with it.

Not Error Tracking
That's Sentry

ScanWarp catches errors and explains them. Sentry gives you full stack traces and regression management.

Not an APM
That's Datadog

ScanWarp measures from the outside. Datadog instruments your code and traces requests through your backend.

Pricing

Free to self-host. Forever.

The full monitoring engine is open source. The cloud version saves you from running infrastructure.

Self-Hosted
$0
forever · MIT license
  • Full monitoring engine
  • All provider integrations
  • AI diagnosis
  • MCP server
  • CLI & notifications
  • Docker Compose setup
View on GitHub
Cloud (Beta)
Soon
currently in private beta
  • Everything in self-hosted
  • No infrastructure to manage
  • Automatic updates
  • Hosted dashboard
  • Global status page
Join Waitlist
Pro (Coming Soon)
TBD
after beta launch
  • Everything in Cloud
  • Multi-region monitoring
  • 30-second check intervals
  • 90-day history
  • Team access
  • Priority support
Join Waitlist

Open Source. Not Open-Washing.

The full monitoring engine, AI diagnosis, MCP server, CLI, and dashboard — all MIT licensed. No feature gates on the free tier. Run it yourself with Docker Compose. Read every line of code.

Star on GitHub

Join the beta waitlist

Get early access to ScanWarp Cloud. We'll notify you when your invite is ready.




Or self-host today — it's free and open source:

CLICK TO COPY $ npx scanwarp init


View Source