Uptime · checked 2026-10-13

Can I vibecode Better Stack?

YES· one-shottable

Checking a list of URLs every minute, alerting when one fails twice in a row and publishing a status page fits in one session, and Uptime Kuma already does it for free. Better Stack's paid value sits elsewhere: probes in several regions that confirm an outage before waking anyone, phone calls that reach you at 3 a.m., on-call rotations for a team, and the logs and traces sold alongside. For a handful of personal sites, a checker on a cheap VPS at another provider covers what matters.

$34/moprice per seat

$408you'd save a year

one sittingbuild time

—people replaced it

Before you start

What you need on hand.

DifficultyIntermediate · Some system setup or keys.

TimeAbout an hour, plus setting up a VPS

Claude Code, Codex or CursorAny of them can run the prompt.
Node.js 22better-sqlite3 compiles a native module, so build tools must be installed.
A small VPS at another providerIf the monitor lives next to your sites, an outage silences it too.
An alert channelSMTP credentials, a Telegram bot token or an ntfy.sh topic.

Before you buildBetter Stack's free plan still watches 10 URLs every 3 minutes with email and Slack alerts. If that covers you, there is nothing to build.

1. Install one agent: Claude Code, Codex, Cursor.
2. Make an empty folder and open it in the agent.
3. Paste the prompt below, press enter, and answer its questions as it builds.

The prompt

Build me a self-hosted uptime monitor with a status page to replace Better Stack
for my own sites. Requirements:

- Node.js 22 + TypeScript, one process: a scheduler plus Fastify serving EJS pages
  and a small JSON API. `npm start` listens on port 4800.
- Monitors live in monitors.yaml (name, type, target, interval, timeout, expected
  status or keyword): HTTP(S) with an optional keyword check, TCP port, and SSL
  certificate expiry with a warning at 14 days. Reload the file without a restart.
- Storage: better-sqlite3 in ./data/uptime.db. Keep every check (time, ok, latency
  in ms, error) for 90 days, then roll it up into daily uptime percentages.
- A monitor goes down only after 2 failed checks in a row, 30 seconds apart, and
  comes back after 1 success. Each down period is stored as an incident.
- Alerts on state changes only (down, recovered, certificate expiring): email via
  nodemailer, plus a Telegram bot or an ntfy.sh topic when set in .env. At most one
  reminder per hour while an incident stays open.
- Heartbeats: GET /ping/:token marks a cron job alive; alert when a ping is later
  than its period plus a grace time.
- /admin (password from .env): every monitor with a status dot, last check, a 24h
  latency sparkline as inline SVG, and open incidents.
- Public /status page: one row per monitor with a 90-day bar of daily uptime and
  the incidents of the last 14 days. No login, no third-party scripts.
- No accounts, no telemetry. Out of scope: phone-call alerts, checks from several
  regions, on-call schedules.
- README: install, a systemd unit, and a warning to run it at a different hosting
  provider from the sites it watches, or one outage takes out both.

Opens your agent with the prompt filled in, you press enter. Or copy it and paste it into an empty folder.

Share on X

Max MiskowSponsoredAI agents, automations and MVPs, built for your company.Visit site ›

Then keep going

Next prompts, same folder.

The first prompt builds the core. Run these one at a time in the same folder to grow it into your own Better Stack.

Add a second probe location to the uptime monitor in this folder. Requirements:

- A small probe mode (`npm run probe`) for another VPS: it reads the same
  monitors.yaml, runs the checks and POSTs results to the main instance,
  signed with PROBE_SECRET from .env.
- The main instance marks a monitor down only when both locations fail within
  the same minute; if only one fails, log it as a regional blip.
- Show per-location latency on /admin.
- If the probe stops reporting for 5 minutes, alert once that the probe is gone.
- Update the README with how to deploy the probe.

Add phone-call alerts to the uptime monitor in this folder. Requirements:

- Monitors tagged critical in monitors.yaml trigger a voice call through the
  Twilio API (account SID, auth token, from and to numbers in .env) when they
  have been down for 5 minutes.
- The call reads the monitor name twice; pressing 1 acknowledges, and no more
  calls happen for that incident.
- Never call more than 3 times per incident.
- Update the README with Twilio setup and a note that each call costs money.

Publish the status page of the uptime monitor in this folder as static files. Requirements:

- Every minute, render /status to ./public-status/index.html plus a
  status.json with the same data.
- Push the folder to a Cloudflare Pages project with wrangler (token in .env),
  only when the content changed.
- The static page shows when it was last updated, so visitors can tell if
  the monitor itself has gone quiet.
- Update the README with the Cloudflare setup steps.

Add a Better Stack importer to the uptime monitor in this folder. Requirements:

- Read a team API token (BETTERSTACK_TOKEN in .env) and page through
  GET https://incidents.betterstack.com/api/v2/monitors.
- Map URL, monitor type, check frequency, expected status codes and keyword to
  entries in monitors.yaml; list anything that has no equivalent.
- Print the new YAML and ask before overwriting the file.
- Update the README with where to create the token.

What you lose

Checks from several regions that confirm a failure before alerting
Phone-call and SMS alerts that escalate to the next person
On-call schedules, rotations and incident timelines for a team
Playwright transaction checks and the logs, traces and error tracking in the same account
A status page hosted away from your own servers

Why people still pay

Teams pay for alerts they can trust at night: a failure confirmed from more than one place, a phone call that escalates when nobody answers, and a status page that stays up when their own servers don't. Once logs and error tracking live in the same account, moving only the monitors saves little.

Moatinfrastructure scale execution polish

Do the math

Better Stack or your own copy?

Seats

Over

Your usage

Count my time3h ×$/h

Better Stack, 1 yr$408

Your own copy, 1 yr$40running + build session

You keep$368pays for itself in the first month

Estimate · A $4–6 VPS runs it, or nothing extra on a server you already have. Email, Telegram and ntfy alerts are free; phone calls through Twilio would be charged per call. The build is token usage on your agent plan.

Leaving Better Stack

Take your data, then cancel.

1 · Take your data out of Better Stack

1Create an API token under Better Stack > API tokens > Team-based tokens, in the Uptime section.
2List your monitors through the Uptime API (GET /api/v2/monitors, paginated) and save the JSON: it holds every URL, interval and setting.
3Status page subscribers come out of GET /api/v2/status-pages/{id}/subscribers, so you can tell them about the new page.

2 · Cancel Better Stack

1Billing can be changed by an admin or a billing admin.
2Lower the responder count under Settings > Usage > Responders and remove paid add-ons such as extra monitors, heartbeats and status-page extras.
3If anything still bills after that, email hello@betterstack.com and ask them to confirm the team is on the free plan; mention the 60-day money-back guarantee if you are inside it.

Changing someone's role to Member does not free their responder license; lower the count yourself.
Members can upgrade themselves to Responder when added to an on-call schedule. Turn that off under Settings > Roles & Permissions > Member, or the bill can come back.
Run the new monitor next to Better Stack for a week before switching anything off, and compare the alerts.

Checked 2026-10-13 against Better Stack's docs: 1, 2, 3, 4, 5

The escape hatch

Don't feel like building it? These are already free.

Uptime KumaOpen sourceSelf-hosted HTTP, TCP, ping, DNS and push monitors with status pages and dozens of notification channels.★ 92k · active Oct 2026 · docker to self-host
WebVisit ↗GatusOpen sourceEndpoints and pass conditions written in one YAML file, with a status page and alerts.★ 12k · active Oct 2026 · docker to self-host
WebVisit ↗UpptimeOpen sourceRuns on GitHub Actions and publishes the status page to GitHub Pages; outages open GitHub issues. No server needed.★ 17k · active Oct 2026 · one-click install
WebVisit ↗

All 8 free alternatives to Better Stack ›

The numbers

Better Stack pricing

Plan	Monthly	Annual, per month	What you get
Free	$0	$0	10 monitors and heartbeats combined, checks every 3 minutes, 1 status page, Slack and email alerts only.
Responder	$34	$29	Unlimited phone-call and SMS alerts, on-call scheduling and escalations, checks as often as every 30 seconds; 10 monitors and 10 heartbeats included.
Enterprise	custom	—	Custom quote: SAML SSO, SCIM, data residency, VPC deployment.

Free tier: 10 monitors and heartbeats in total, 3-minute checks, 1 status page, no phone or SMS alerts.
Billing: Monthly, or yearly at about two months off ($29 instead of $34 per responder). A 60-day money-back guarantee is advertised.
Hidden costs: Most extras are priced one by one: 50 more monitors cost $25 a month, 10 more heartbeats $20, each extra status page $15, password protection $50 per page and white-labelling $250 per page. Admins and team leads use a responder license too, and changing someone's role does not lower the license count by itself.
Source: betterstack.com

Also on the list

Cronitor$15/mo · Cron monitoringYES

Vercel$20/mo · HostingKINDA

Airtable$24/mo · DatabasesYES

Questions

Better Stack, briefly

Not answered here? Tell us and we'll add it.

Ask a question ›

Free watches 10 monitors or heartbeats every 3 minutes, with one status page and Slack or email alerts.

Yes. An AI coding agent such as Claude Code, Codex or Cursor can build a usable personal version in one session with the prompt on this page, running on your own machine.

About $34 a month on the Responder license plan per seat, $408 a year (checked 2026-10-13).

Checks from several regions that confirm a failure before alerting; Phone-call and SMS alerts that escalate to the next person; On-call schedules, rotations and incident timelines for a team; Playwright transaction checks and the logs, traces and error tracking in the same account; A status page hosted away from your own servers. If any of those matter to you, keep paying.