What's newin FastMonitor.

• Read guides and product notes at the blog: cron job monitoring, SSL and DNS checks, alert routing, Slack and PagerDuty, status pages, the agent plugins and the telemetry API.
• Browse by category, and follow new posts through the RSS feed.

• Monitors: websites and HTTP APIs with assertions, TCP ports, SSL certificates, DNS records, domain expiry, broken links, and cron jobs or workers that report each run. Each check can wait for a number of failed checks in a row before it alerts anyone.
• Alerts: email, Slack, signed webhooks and PagerDuty, depending on your plan, sent to notification lists. On Enterprise, routing rules pick the lists, and "Where would this alert go?" previews a rule's effect.
• Status pages and reports: public status pages on Starter and Enterprise, and scheduled email reports on every plan.
• Teams and security: organizations with owner, admin, billing, member and viewer roles, two-factor sign-in with backup codes, and Google or GitHub sign-in. The dashboard speaks 8 languages.
• API and agents: the telemetry API for job runs, and an MCP endpoint for coding agents on every plan.
• The docs now cover webhooks and signature checks, HTTP assertions, SSL, DNS and link checks, notification lists, status pages, routing, automation, teams and API keys. See every plan on Features.

• The dashboard now downloads only the page you open. Its largest download went from about 2.5 MB to under 400 kB.
• Only your language is downloaded; the other seven load when you switch.
• Lists stop refreshing while the tab is in the background and catch up as soon as you come back.
• Monitor lists load 50 monitors at a time, with Load more for the rest.
• A public status page no longer downloads any dashboard code.

The last hard-coded English labels in the dashboard now go through translations, and missing translations were filled in for all eight languages.

• No page on the website scrolls sideways on a phone or tablet, and the menu opens and closes on every page.
• Images are smaller and served as AVIF or WebP where your browser supports them, so pages load faster on mobile connections.
• The website supports Safari and iOS 16.4 and later.

• New Features, Integrations, API and FAQs pages describe what FastMonitor does today, with plan tags where a feature is not on every plan.
• The routing, cron job, automation and AI sections on the home page show the real product, and pause their animations when you prefer reduced motion.
• Placeholder copy is gone from the product pages.

Public status pages are rendered on the server and cached, so they open quickly even for visitors arriving from an incident link. Uptime figures refresh every 30 seconds; current status and incidents are always live.

• Checks run side by side, so one slow site no longer delays the checks behind it.
• Broken-link crawls that hit their time limit now report the pages they checked instead of a timeout.
• Alerts are sent in parallel, and scheduled report emails retry with backoff and stop after a fixed number of attempts.
• Old check results are cleaned up daily outside peak hours, following your plan's retention.

Every API response carries an X-Request-Id header, and server errors include the same requestId in the response body. Quote it when you contact support and we can find the exact request.

Ask Claude Code, Codex, Hermes Agent or any MCP client what is failing, triage an incident or plan a check.

• Settings → AI agents & MCP creates MCP tokens with only the permissions you tick and an expiry, shows each token once with ready-made setup for each agent, and lets you revoke it at any time.
• FastMonitor serves a remote MCP endpoint at your API address followed by /mcp, with five tools: list monitors, monitor summary, open incidents, incident summary and acknowledge incident.
• Tokens see only your own monitors and incidents, are rate limited and every call is audited. See the docs.

• The Failing filter shows every failing monitor, including jobs that missed a run; Healthy no longer includes failing or paused monitors; and there is a new Paused filter.
• If a list fails to load, you see the error with Retry instead of an empty list.
• A check's details show only the settings its check type uses, with the real tolerance and reminder values.
• Monitor cards open with the keyboard, and errors show the actual reason instead of [object Object].

• If a check fails to load for editing, you see an error with Retry instead of an empty form, so saving can no longer create a duplicate check.
• Validation errors appear next to the field they belong to.

• A job's time zone and platform stay exactly as you set them when you edit and save it. Jobs in UTC, Japan or US Eastern time are no longer moved to London time.
• The dashboard and the API accept the same cron expressions, and invalid fields get a clear message instead of a server error.
• Alert options, notes and follow-up reminders are saved when you create a job as well as when you edit it, and a grace period of 0 is kept.
• "Next in" counts down to the run FastMonitor actually expects.

• New email sign-ups confirm their address before creating monitors.
• Two-factor codes are rate limited per account and per network address.
• Signing in with a social account no longer places a token in any URL.
• The telemetry API answers 429 RATE_LIMITED when a key sends too many requests; retry with backoff.
• Behind the scenes: stricter production secrets, reproducible database migrations and automated checks on every change.

• Telemetry events are always recorded against the monitor your API key is allowed to report for.
• The Python and PHP SDKs use the documented POST /e/execution-events call; their older ping helpers still work and now warn that they are deprecated.