Error detection
Server error detection catch what a 200 hides
PHP warnings, WordPress critical errors, SQL errors, and stack traces, detected on the pages you already monitor. Nothing to install, nothing to configure.
# last scan · agency-client.example
2 findingsPHP error output visible on page
PHPDeprecated: preg_match_all(): Passing null to parameter #2 ($subject) is deprecated in /home/…/wp-content/plugins/…/Shortcodes.php on line 525
PHP error output visible on page
PHPWarning: Undefined array key "title" in /home/…/wp-content/themes/…/header.php on line 12
The page returned 200 OK. Every uptime check passed. The deprecation spew was stacked above the header, where every visitor could read it.
Why error detection
The site was up. The page was broken.
A PHP version bump or a plugin update flips one setting, and suddenly every visitor sees deprecation warnings stacked above your header. The server returns 200, so your uptime check passes, your host's monitoring passes, and nobody tells you. Sentinel reads the page the way a visitor does and recognizes the shape of leaked server errors on any stack.
- Catches error output that renders on an otherwise healthy page
- No configuration: the signature library ships with Sentinel
- Included on every plan, free plan included
Error detection settings
- Detection
- On by default
- Stacks
- PHP · WP · SQL · Python · .NET +
- Matching
- Structural (path + line)
- On detection
- Warning incident + alert
Features
Detection that ships configured
Zero configuration
On by default for new monitors. No error strings to guess, no rules to write. Sentinel already knows what leaked server errors look like.
Any stack
PHP, WordPress, Craft CMS, Statamic, Laravel, Symfony, Python tracebacks, ASP.NET, and SQL errors. One detection engine, one signature library.
Structural matching
Signatures require the shape of a real leak: a file path plus a line number, an SQLSTATE code, a traceback header. A blog post about PHP errors never alerts.
A security signal too
Visible error output leaks server paths, and a debug page in production is an exposure. The alert tells you what leaked so you can judge the urgency.
Warning, not panic
The site is up, so this opens a warning-severity incident rather than paging you like an outage. You get told before your clients do.
Resolves itself
Fix the page and the incident closes on the next scan, with a recovery alert. Full incident history, timestamps, and the exact text that was found.
By the numbers
The check nobody else runs
External uptime monitors check status codes and leave you to hand-write keyword rules for error strings you have not seen yet. Sentinel ships the signatures, scans the pages it already checks, and turns leaks into incidents.
10+
Signature families across stacks
0
Rules you have to write
24/7
Body scanning on pages you already monitor
100%
Of plans include it, free included
Which stacks and frameworks does it detect?
PHP errors and stack traces (which covers WordPress, Craft CMS, Statamic, and any PHP site), the WordPress critical-error and database-connection pages, SQL errors like SQLSTATE codes, Python tracebacks, Laravel and Symfony debug pages, and ASP.NET server error pages. The signature library grows as we find new shapes worth catching.
Will it alert on a page that just talks about errors?
No. Every signature requires the structure of a real leak, such as a file path with a line number or an SQLSTATE code, not just the word 'error'. A tutorial about fixing PHP warnings will not trigger it.
My host already monitors my site. Why do I need this?
Host monitoring watches the server, from inside the host's own network, for the host's own SLA. Error text rendering on a 200 page is invisible to it. Sentinel checks the page from outside, the way your visitors experience it.
Is it enabled on my existing monitors?
New monitors have it on by default. Existing monitors keep it off until you enable it in the monitor's settings, so adding the feature never opens incidents for warnings a site has quietly carried for months. Turning it on takes one checkbox.
Which plans include server error detection?
Every plan, including the free plan. Alerts go through your configured channels: email and Discord on the free plan, plus Slack, Teams, SMS, and webhooks on paid plans.
What happens when the error is fixed?
The incident resolves automatically on the next clean scan and Sentinel sends a recovery alert. You keep the full history: when the leak appeared, what it said, and when it cleared.
Your visitors can see that stack trace
Add a monitor and error detection is already on. Find the leaks before a client screenshots them.
No card for the free plan · 14-day trial on paid plans · Cancel anytime