Changelog

Improvements

• Updated the Lighthouse checker to correctly audit HTTP sites when requested. This avoids automatic upgrades to HTTPS caused by modern browsers and ensures more accurate results.

• Significantly improved marketing site performance:

  • Converted over 100 SVGs to sprites for faster rendering.
  • Reduced overall JavaScript bundle size.
  • Removed unused libraries from marketing pages, including Chart.js.
  • Authentication-related scripts now load only on login and registration pages.
  • Replaced Axios with the browser’s native fetch() API.
  • Removed the Gill Sans webfont, reducing unnecessary font loading.
  • Applied multiple frontend optimizations to improve load speed and responsiveness.

• Improved menu performance by only initializing interactive components when needed, reducing page-load overhead.

• Enhanced blog performance and SEO:

  • Fixed Open Graph metadata for blog posts.
  • Improved image loading with preloading and prefetching.
  • Added missing author, timestamp, and alt-text metadata.
  • Ensured SEO tags are consistently passed to the main blog views.

• Improved user seeding logic so new Oh Dear users always land on the Oh Dear team after login, instead of a reseller team.

Fixes

• Fixed Lighthouse checks incorrectly failing when disabled metrics or categories exceeded thresholds.
• Fixed a metadata issue where the monitor name was used instead of the site name in blog Open Graph data.

New

• Added support for mobile Lighthouse checks, allowing performance audits to better reflect mobile user experiences.

Improvements

• Improved sitemap URL reachability to ensure more reliable sitemap checks.
• The Public API now lets you set the new_links_only and force_crawl_urls broken link settings when creating or updating monitors.

Fixes

• Fixed an issue where cron titles could overflow the page width in the monthly report UI.

New

• Added Grok support to the AI checker.

Improvements

• Oh Dear is now compatible with PHP 8.5.
• The AI checker now shows a cost warning when selecting an hourly check frequency.
• Updated documentation for clarity and accuracy.

Fixes

• Preserved query strings in Status Page history and secret update links.
• Fixed Status Page history links to consistently use relative URLs.
• Fixed an issue in the Status Page template editor where typed text did not fully appear while editing.
• Fixed a broken link in the “How to configure maintenance windows” documentation.

New

• Added an anomaly detection Slack alert to flag potential worldwide issues when an unusually high number of notifications are being sent.
• Added new Claude models to the AI checker.
• Launched a new Careers section on the site.

Improvements

• Improved the AI checker browser session to return clear JSON validation errors instead of redirecting.
• Truncated large AI checker JSON outputs to keep results readable; users are guided to screenshots for visual verification when needed.
• Fine-tuned AI checker prompts to reduce token usage.
• Reduced the maximum token limit to 30,000.
• Cleaned up AI checker HTML output by trimming unnecessary SVGs, images, and extra markup.

Fixes

• Fixed a Passkeys issue caused by a missing configuration value after a package upgrade.
• Performance response time is now correctly included in the Monitors API response.

Improvements

• Simplified how VAT is handled in Stripe / Cashier.
• Enabled caching on status badges. Previously, when session middleware was active, Cloudflare ignored cache-control headers due to set-cookie behavior; this is now resolved.
• The used-ips endpoint is now cached through Cloudflare for improved performance.
• Sitemap payloads are now always encoded as UTF-8 to ensure correct display in notifications.
• Optimised sitemap generation memory usage to avoid OOM issues.

Fixes

• Corrected Japanese translations on status pages.
• Custom headers are now correctly sent during scheduled sitemap runs.
• Fixed a UI issue in the monthly uptime report email.
• Resolved a problem where the performance summary always returned “Error.”
• Request Timeline: fixed errors when totalTime was 0 and corrected percentage calculations.
• Lighthouse notification settings no longer overwrite frequency_* values.
• Fixed rounding issues for Lighthouse metric values in both notifications and the UI.
• Updated default metric thresholds for Lighthouse notifications.
• Prevented notifications from sending during an active maintenance period.
• Resolved false positive “problem fixed” events for skipped health checks
• The crawler now correctly reports URLs with new lines in the HTTP Host header as invalid.

New

• Added longer notification snooze options, including 6 months and 1 year.

Improvements

• Our robots.txt crawler now sends the correct headers.
• Disabling the scheduled task check no longer deletes its cron definitions.
• Application Health checks no longer validate SSL certificates, since SSL validation is already handled by the Certificate Health check.

Fixes

• Fixed a broken toggle in the DNS monitor settings.
• Fixed a server error that occurred when creating a Status Page update template with a very long title.

New

• Added support for multiple Lighthouse servers per continent to improve performance and regional accuracy.
• Improved compatibility with password managers by adding a well-known passkey endpoint.
• Added a Slack alert that triggers whenever the peak monitor count is updated.

Improvements

• Lighthouse checks are now dispatched hourly instead of daily. The scheduler now correctly determines whether a Lighthouse check has run recently, ensuring checks aren’t skipped when queue processing takes longer than expected.

New

• Added an option for resellers to allow managed team users to create, update, and delete monitors.
• The Status page now shows the full update history, not just the last 7 days.

Improvements

• Updated reseller documentation to clearly explain the different available user roles.

Fixes

• Fixed an issue where CLS input fields in Lighthouse settings did not accept decimal values.
• Adjusted sorting on the Broken Links report so ignored links now appear at the bottom of the list, making new or active issues easier to find.

New

• Added a request waterfall timeline to uptime checks, providing clearer insight into each stage of request execution.
• Added AI monitoring to the Oh Dear documentation features page.
• Introduced a feature allowing resellers to remove managed teams.

Improvements

• Improved performance records ingestion, resulting in faster data display across the UI.
• Updated copy across multiple sections: DNS, Broken Links, Lighthouse, and Uptime Monitoring pages, for clarity and consistency.
• On the Broken Links report UI, newly discovered links now appear at the top of the list, ensuring they are prioritized over ignored ones.
• Added a JavaScript client example to the MCP server documentation.

Bug Fixes

• Fixed an issue that broke the UI when updating notification throttling and frequency settings.
• Fixed an issue where, if two monitors went down simultaneously on the status page, only one monitor’s updates were displayed.

New

• Teams can now choose custom plans, providing more flexibility in managing subscriptions.

Improvements

• Improved uptime charts and the preview charts on the monitor list. Charts now correctly display when uptime data includes sub-millisecond values, instead of appearing as a flat line.
• Enhanced documentation for clarity.

New

• Added support for Stripe metered billing for resellers.
• Skipped non-unique URL checks in image sitemaps.

Improvements

• Refactored scheduled task logs:
  • Logs are now retained for 45 days.
  • At least 10 pingFinished entries are always kept.
  • Better handling of pingStarting/pingFinished pairs.
• Improved scheduled task validation, only form fields relevant to the selected cron type are now validated.
• Refined UI across multiple sections for enhanced usability and visual consistency.
• Ensured all documentation categories are now included in the mobile menu.
• Added average packet loss to the uptime chart tooltip and fixed sub-millisecond display precision.

Bugfixes

• Fixed an issue where the secondary location tab showed incorrect HTTP response headers.
• Added unique wire keys to prevent input leakage on Uptime, Notifications, and Broken Links settings pages.

New

• Added a new UI section showcasing all our third-party integrations, everything you can connect to Oh Dear in one place.
• Added a visible meta payload in scheduled task details so users can easily see when there was an issue during execution.

Improvements

• The uptime secondary checker now runs from the same continent as the primary one, improving accuracy for users with regional geo-blocking and ensuring compliance for EU-based users who require checks to stay within their region.
• Added an “AI setup” option in the monitor creation flow. It’s currently in closed beta, with a “Coming soon” notice shown for all other users.

Improvements

• Improved UI on the monitor list, removing the small chart stutter that happened during poll refreshes.
• Polished the uptime settings interface for a cleaner experience.
• Uptime buckets on the status page are now timezone-aware.
• Status page queries have been optimized to load much faster.

Fixes

• Fixed an issue in the Crawler for URLs that contained .. (dot-dot) syntax.
• Fixed minor styling inconsistencies across the blog.
• Fixed an issue with the throttle_failed_interval_unit in Notification Settings that caused incorrect throttling behavior for some users.

Improvements

• in the uptime results we now show results from both uptime checkers, should the first one fail
• on the billing screen, you can now see your invoices without having to go to the Stripe portal

• added a OpenAPI spec to our API docs
• performance improvements
• fix bugs around whitelisting broken URLs
• building AI monitoring
• launched Request Mirror
• lauched DNS info

Improvements

Big performance improvements to our website, dashboard and uptime checks. more details

Major new features

• We've added 🏓 ping and 🔌 TCP Port monitoring. See: Introducing Ping & TCP port monitoring
• you can now customize the frequency of the checks more details

Updates

• we've published a new major version of our SDK

Gearing up for a big feature update that introduces Ping & TCP monitoring early next week 👀

This week, we introduced an option to prevent a status page from being index by search engines.

Meanwhile, we're also hard at work polishing the upcoming Ping / TCP monitor. We're now testing this extensively in our staging environment.

Improvements

• we've added notification logs: https://ohdear.app/news-and-updates/introducing-our-new-notification-logs

Meanwhile we're also focussing on launching our Ping / TCP checks

Improvements

• we've added a new uptime check location: Stockholm 🇸🇪

Improvements

• we are working on adding support for ping and TCP monitoring. You'll find some more info in this blog post.
• we slightly adjusted the layout of the app: we now use Inter as the base font.

Fixes

• we'll now deleted configured scheduled jobs monitors when the cron check is disabled

Improvements

• a new "Detected Certificates" tab was added to the certificate health check screen. This tab will show all certificates we dedected for a site.
• we added a troubleshooting screen for lighthouse checks
• add graph was adding to the lighthouse result history
• on the new "Usage" screen you can now see how much sites / scheduled jobs you are using against your plan limits
• for the PHP developers amongst you: you'll be happy to know our code base has reached PHPStan level 6 💪
• show application health client error deails in application health report alert message
• add named windows to 'start-maintenance' api endpoint
• add Cloudflare email script link to global broken links whitelist
• complete static analysis coverage for the entire application

Fixes

• small styling fixes
• fixed the 'domain' section in monthly reports for sites with unsupported RDAP TLDs
• fix seo meta tags for docs
• render correct site count when 'sites with issues' tab selected
• fix internal errors when handling partial payments

Improvements

• you can now name a maintaince period, both in our UI and via our API
• we've improved the perceived performance of the list of sites: we first render placeholders, so the users get a fast response, and then replace those with the actual data as soon as it is loaded.
• we can now notify you via Google Chat
• we now accept payments via Revolut
• our registration page has been streamlined
• we now mention our available payment methods on the pricing page and in the FAQ
• our internal workflow to publish the changelog you are now reading has been improved

Fixes

• ensure it is clear which things are clickable, by adding the cursor-pointer CSS class to those elements
• all broken links on the marketing website are fixed
• our crawler checks are scheduled more reliably
• we now correctly clean up old historical data

Improvements

• add additional (and standardise) url validation
• move uptime summary below uptime chart

Internal

• continuous improvements to static analysis coverage
• bug fixes

Fixes

• apply all custom headers to application health check
• apply correct host header checks for status pages and GTM scripts

Improvements

• restored uptime summary component in uptime/performance results
• add retries and delays when calling third-party APIs during monitoring
• increased static analysis level for internal testing

Fixes

• fix run result alerts in multiple check components
• clarify run result errors in multiple check components
• fixed performance chart scaling issues
• improved exception handling in checks and reports
• other bug fixes

Improvements

• updates to uptime redesign
• updates to check report page alert cards

Fixes

• fixed a an issue where performance metrics were sometimes not being stored for offline sites

Improvements

• automatically disable lighthouse reports that error (or time out) three times in a row an alert the team owner
• streamlined onboarding for new teams making it easier and faster to create your first site

Fixes

• added clearer alert text to distinguish between SSL certificate and domain-related subjects
• only use Google Tag Manager for Oh Dear domains on status pages

Internals

• removed backend dependencies
• fix dependency and build warnings for front-end assets
• the team is making final touches to an uptime and performance redesign (coming soon!)

Improvements

• redesigned and simplified Oh Dear blog
• added monthly report settings to API
• added tag filters to site API

Fixes

• added x.com to list of crawler ingore list defaults
• fixed API notification permissions issues

Improvements

• added support for passkey authentication
• added API support for snoozing to individual cron and health check items
• added support for monthly site settings via API
• internal admin updates to help customer support create custom pricing for enterprise users
• internal admin updates tracking cancellation reasons to understand customer usage
• removed unused team check history (site check history is still available)

Fixes

• fixed an issue with login via passkey authentication
• cleaned up site snooze history UI to clearly show active tabs

Improvements

• added performance charts and tables to monthly site report
• added lighthouse report summary to monthly site report
• internal reporting, guides and docs to help with support

Fixes

• fixed 'ignore' broken link issue with long URLs

Improvements

• Added changelog RSS feed
• Speed improvements to dashboard and performance graphs

Behind the scenes

• Removed all code related to new ping endpoint
• Fixed an issue with deleted site data not executing on schedule
• Fixed Clickhouse managed database issues (compute and sharding)
• Optimised Clickhouse ingress and queries for performance reporting

New features

• On the snooze screen of a check, we now display the history when the check has been snoozed
• We've activated new onboarding and offboarding email flows

Behind the scenes

• All of the uptime and performance statistics have been moved from MySQL to ClickHouse, unlocking future improvements
• By leveraging parallel testing, we made our test suite run 3x faster on GitHub Actions

Fixes

• Fixed a bug where times were not displayed in the correct timezone on the scheduled jobs log
• We’ve added help text on various notification screens, making them easier to use
• Our status page now gets automatically updated, even if the uptime check is snoozed
• We’ve improved the error handling of our DNS checker, making it more reliable
• The sitemap checker can now correctly process empty sitemaps
• We increased the maximum run time of the sitemap checkers, enabling checking larger sitemaps
• A bug was fixed that prevented status pages from being automatically refreshed

Improvements

🐘 We performed a couple of upgrades and we're now running our entire service on PHP 8.4 and Laravel 12 🥳

Improvements

• When a site is considered offline because of too many redirects, we add a visual indicator to the UI that shows where in the redirect process the limit was reached
• Add datetime tooltips for uptime & scheduled tasks, to more clearly show UTC/server/user timezones

Fixes

• Bugfix timestamp for scheduled task notifications not properly showing
• Allow longer application health check URLs on input validation

Internals

• Better insights & notifications to subscription updates

New features

• Allow toggling to only receive new broken links notifications (only when we detect that there are new broken links compared to the previous run)

Fixes

• Fixed some missing translations on certificate detail pages

Improvements

• New and improved ping.ohdear.app is now processing incoming traffic, around 5x faster than the old system 🎉

Fixes

• Fixed performance chart on dashboard rendering backwards in edge cases
• Bugfix custom thresholds for certificate expiration monitoring

Improvements

• Extended the Broken Links detail pages with fine-grained "Solutions", offering insights & actionable tips to resolve the detected problems
• Added ability to override robots.txt when crawling, allowing "force crawl" and "force do not crawl" URL options
• Tweaked blog layout for minor improvements
• Better internal management of our testimonials
• Applied response-caching for all marketing pages of the site (server-side rendering from 500ms -> 100ms)

Fixes

• Applied smarter rate limiting & retry logic when deliver bulk notifications to the same destination (ie: same Slack channels)
• Fixed dark-mode styling on the homepage for small top-banner missing CSS
• Fixed performance issue on uptime check pages not loading when sites were down for > 24hrs
• Fixed the "Feature announcements" posts on the blog not sorting properly

Improvements

• Automated deployments to staging & production for new ping-beta endpoint
• Updated layout for the blog
• Finished changes to the laravel-schedule-monitor package to include a default API URL and optional different ping endpoint
• Added caching to our Status Pages to better handle surges in traffic
• Updated Telegram Notification setup instructions for clarity

Fixes

• Fixed wrong IP for the New York Lighthouse server being reported in the IPs Oh Dear uses

Improvements

• Fixed message when certificates expire, to show actual time remaining before expiry instead of "soon"
• Affiliate page updated
• Introduction of this Changelog section 🥳
• Internal improvements to our helpdesk/support integrations
• Clarified the setup instructions for Telegram Notifications

Improvements

• Pass-through IP address for ping-beta customers
• Allow local testing of ping callbacks
• Show Sitemap and Sitemap URL issues separately
• Filter invalid Markdown tags when configuring a Status Page
• Replace bespoke cookie banner with Cookiebot

Fixes

• Prevent status update requests as a user is typing

Improvements

• Allow a sites' Friendly Name to be configured straight in the Add Site modal
• Ensure the check names are always human readable
• Expose Sitemap crawling issues more clearly to the end-user
• Add API documentation for creating status pages & status page sites
• Upgraded from Tailwind CSS 3 to Tailwind CSS 4

Beta releases

• A newly architected ping architecture for receiving cron jobs/scheduled tasks pingbacks is ready for internal testing

Activity

• The team is attending LaraconEU 2025 to meet with clients & like-minded developers 🥳

Improvements

• Redesigned the "Add Site" screen for simplicity, advanced options are now hidden by default
• Add download and uptime exports per site
• Add daily downtime summary export
• Add sync status page sites api endpoint
• Add create status page api endpoint
• Add delete status page site api endpoint

Fixes

• Fix for handling of bounced e-mail notifications