Skip to content
UpButler
ProductAgentsStatus pagesDocsPricing
Sign inStart free

Changelog

What we shipped, newest first. Follow along on our status page or ask your agent to read this page.

Plans

New plans: Starter, Pro and Business

A cheaper first step for a page without our branding, a Pro plan priced like the rest of the market, and Business for teams on call. Every new workspace starts with 14 days of Pro.

  1. Four plans

    Free, Starter at $12/month, Pro at $29/month and Business at $99/month; yearly is ten times the monthly price. Business replaces Team and adds the audit log viewer, 20 browser checks and 500 components per page.

  2. Starter: your page without our branding

    25 monitors every minute from 2 regions, 500 email subscribers, 1 custom domain and no "Powered by UpButler" line.

  3. 14 days of Pro for every new workspace

    No card. When the trial ends the workspace moves to Free unless you pick a plan, and nothing is deleted: extra monitors are paused and extra components hidden until you upgrade. Workspaces created by an agent start the trial when a person claims them.

  4. Free, refocused

    10 monitors every 5 min, 10 components and 25 email subscribers per page, 10 AI reports and 300 emails a month. Commercial use stays allowed, and the API, MCP server and webhooks stay unmetered on every plan.

  5. Enterprise on request

    Custom limits and regions, a DPA, an SLA and annual invoicing, from $299/month. Write to [email protected].

Compare plans

v2.1

v2.1: on-call, regions and deploys

The things teams asked for once v2 was running their status pages: someone on call, checks from more than one place, and a straight answer to "was it the deploy?".

  1. On-call rotations and escalation

    Weekly rotations with overrides, and escalation policies that page the person on call, specific members or any alert channel, level by level, until someone acknowledges. Paging goes out by email and Telegram (no SMS or phone calls). On-call is part of Team (now Business); acknowledging incidents, which stops reminders and escalation, works on every plan, from the dashboard, the API or a link in the alert.

    On-call docsEscalation policies

  2. Multi-region checks with quorum

    HTTP, TCP, DNS and manifest monitors can be checked from Helsinki, Nuremberg, Falkenstein and Lauterbourg in the same round. A monitor goes down only when a strict majority of the reporting regions agree; an even split is degraded, a failing minority stays visible per region. Probes that stop reporting are left out instead of counted as failures. 1 region on Free, up to 3 on Pro, all of them on Team.

    Multi-region checksHow quorum works

  3. Deploy markers and AI deploy correlation

    Record releases from the API, the CLI, an agent or the new GitHub Action. Deploys show up on monitor charts and incident timelines, and AI incident reports now say when errors began right after a release.

    Deploys docsGitHub Action

  4. Account security

    Sign in with Google or GitHub as well as email. TOTP two-factor with recovery codes, and owners can require 2FA for the whole workspace. See and sign out your sessions, and read an audit log of sign-ins, key, member, billing and domain changes, kept for 400 days.

    Security docsOur security page

  5. AI that writes in your voice

    Each status page can carry AI guidelines and a list of banned terms; AI public text that breaks them falls back to a neutral template. New: a weekly digest of your monitoring week (email for members who opt in, plus a digest.weekly event), AI postmortem drafts you can edit and publish to status pages, and incident templates with placeholders.

    Digest, postmortems and templates

  6. Blip protection for pushed components

    Pushed and manifest statuses that are worse than the current one now need two fresh reports before they apply (configurable per component); recoveries still apply at once. Senders that already confirmed can say so with confirmed: true.

    Push and manifest

  7. Importers

    Bring monitors, heartbeats and status pages over from UptimeRobot and Better Stack (plus alert contacts from UptimeRobot), and components, groups and incident history from Atlassian Statuspage (a public URL is enough). Every import starts with a dry run, never stores your credentials, and can be re-run safely.

    Import docs

  8. A livelier dashboard

    The dashboard now updates live as checks come in, and new workspaces start with an onboarding wizard.

  9. CLI and SDKs

    The upbutler CLI covers monitors, heartbeats (including wrapping a cron command), incidents, deploys, MCP setup and any API operation, with --json for agents. SDKs for JavaScript and Python include webhook verification.

    CLISDKs

v2.0

UpButler v2: built for agents

A rewrite of everything: a new API, new checks, new status pages, and a product that AI agents can operate end to end. Humans get the same power in a faster dashboard. Here is what is new.

  1. Agents can sign up on their own

    POST /api/v1/agent/bootstrap returns an API key and a claim link. The agent starts working immediately; a human claims the workspace within 7 days to keep it and unlock email.

  2. MCP server

    https://upbutler.com/mcp exposes 50+ tools over streamable HTTP: monitors, status pages, components, incidents, maintenance, subscribers, channels and the event stream. Every tool is the same operation as the REST endpoint.

  3. services_add: one call per service

    Creates a monitor and the status page component that follows it, creates the page if needed, and returns ping URLs for heartbeats.

  4. Manifest monitors

    One JSON endpoint drives many components. Accepts flat maps, arrays and health-check formats, and reads Statuspage components.json as is.

  5. Push statuses in batches

    POST /pages/:id/status updates many components in one request, with optional messages and a TTL that marks silent components unknown.

  6. AI incident reports

    Likely causes, suggested actions and a customer-safe public summary when an incident opens, a post-incident summary on recovery, and "polish with AI" for your own notes.

  7. Themeable status pages

    Theme presets (Midnight, Grid, Daylight, Paper), self-hosted fonts, accent colors, logos, groups and 90-day uptime bars. Custom domains and custom CSS on Pro.

  8. Subscribers on every channel

    Email with double opt-in, webhooks with a challenge handshake, Slack, Discord, RSS and Atom. Subscribers can follow individual components.

  9. Signed webhooks

    Every outgoing webhook is signed per the Standard Webhooks spec, so you can verify it with any of their libraries.

  10. Quieter, smarter alerting

    Confirmation re-checks, failure and recovery thresholds, failures grouped into one incident per page within 30 minutes, and reminders at 30 minutes, 2 hours and 8 hours.

  11. New check kinds

    JSON, header, latency and TLS-expiry assertions on HTTP checks; DNS checks; scripted API checks on Pro; real-browser Playwright checks on Team (now Business).

  12. New plans

    Free (10 monitors, 30 AI reports), Pro at $15/month (50 monitors at 30 s, custom domains) and Team at $49/month (250 monitors, browser checks, 25 members). Billing through Polar. Replaced by the October 2026 plans above.

Try the new beginning, from any agent:

Terminal
curl -X POST https://upbutler.com/api/v1/agent/bootstrap \
  -H "Content-Type: application/json" \
  -d '{"agentName":"Deploy bot"}'

Start with the agent quickstart or the MCP guide.

v1

The original UpButler

The first version: HTTP monitors with scripted tests, simple public status pages, incidents and notification channels. It taught us what people actually needed from monitoring, and most of all that the next user would often be an agent. Everything from v1 lives on, rebuilt, in v2.