Page is loading...
Office hours: Mon–Thu 9 am–5 pm Fri 9 am–3 pm (CET) hello@hmf-it.de

A hacked plugin, one weekend — what happens in a real WordPress security incident

A case study from Osnabrück, Germany: how the clean-up went, how these attacks work technically — and how to tell whether your own website is affected.

Friday, late afternoon

A while ago, a security vulnerability in a widely used WordPress plugin was being actively exploited. One that was also running on websites I look after.

News like this never arrives on a Tuesday at ten. It arrives on a Friday, late in the afternoon.

If your first question right now is “How would I even know whether my WordPress website has been hacked?” — that is exactly what this post is about. I will take you through a real incident: what happened, how the clean-up went and what you can take away from it for your own website.

What happened next

The weekend did not go as planned: every affected website checked, evidence preserved, cleaned up. The plugin developer reacted admirably fast and released a patched version at short notice — I installed it everywhere straight away and renewed all credentials and security keys on top. By Sunday evening the last site was done.

On Monday every client had a clear report: what was checked, what was found, what was done. That report was the first any of them heard of the incident — not a broken website, not a Google warning, not a legal warning letter (an “Abmahnung”, the formal cease-and-desist letter common in Germany).

How an attack like this works technically

Without going too deep into the details of this particular case (the vulnerability has since been closed), the pattern is almost always the same — and it helps to know it:

  1. The vulnerability becomes public. A security researcher or the developer publishes the weakness. From that moment on, a race is under way.
  2. Automated scanners swarm out. Within hours, attackers work through millions of websites checking whether the vulnerable plugin is installed. Your website is not a personal target — it is simply an IP address on a list.
  3. The malware moves in. Through the gap, a so-called web shell is planted — a small file that lets the attacker control the website remotely. Typically this comes with: a secretly created administrator account with an inconspicuous name, hidden backdoors in system directories and tampered plugin files. All of it designed to survive an update.
  4. The actual exploitation comes later. Sending spam, redirecting visitors to scam sites, extortion, or simply selling the access on. Weeks often pass between the break-in and any visible damage.

And this very pattern is repeating itself right now: just this week, the BSI (Germany's Federal Office for Information Security) warned about the “wp2shell” vulnerability chain in WordPress itself — risk rating “critical”, active attacks within hours of disclosure. The time between a vulnerability being published and being abused shrinks year after year. If you only hear about a Friday vulnerability on Monday, chances are you are already too late.

How to tell that your website might be affected

A few warning signs you can check even without specialist technical knowledge:

  • Unknown administrators: Look under “Users” for accounts you did not create. Attackers choose inconspicuous names and back-date the creation date.
  • PHP files in the uploads folder: The wp-content/uploads directory is for images and documents — not .php files. Any PHP file in there is suspicious.
  • Google warnings: Messages in Search Console (“This site may be hacked”) or rankings that suddenly collapse.
  • Strange redirects: Visitors end up on other people's sites, but only sometimes — attackers often redirect only visitors arriving from search engines so that the site owner notices nothing.
  • Changed behaviour: The website becomes slow, sends emails nobody wrote, or your hosting provider reports unusual traffic.

Important

The absence of these symptoms is no proof of a clean website. Good attackers are invisible — which is why, if you suspect something, you need a real inspection of the files and log files.

Three things I take away from the incident

1. The vulnerability is rarely the problem. The reaction time is.

Security vulnerabilities in plugins are everyday business — new ones are published every week. What matters is how quickly somebody hears about them and acts. A website with nobody responsible for it has nobody to react in an emergency. The difference between “patched on Friday evening” and “sometime next week” is the difference between a 20-minute update and a weekend of forensics.

2. “Updates run automatically” is not a security strategy.

Automatic updates are good and proper — but they do not check whether something already happened before the update. If all you do is update, in the worst case you seal the malware in: the gap is closed, the backdoor stays inside. That is why every actively exploited vulnerability calls for a compromise assessment — comparing the files against the originals, a look at the user list, a look at the server logs.

3. Every plugin is a flatmate.

It lives in your website, has a key — and brings its own weaknesses with it. The most honest security measure costs nothing: clear out regularly. Every deactivated, orphaned or “we might need that again some day” plugin enlarges the attack surface. Every line of code that is not there cannot be attacked.

The uncomfortable question

If a plugin on your website became a target tonight: who would notice tomorrow morning?

If the honest answer is “nobody”, that is not an accusation — but it is a good reason to change it. Whether with me or with somebody else: your website needs a person who is responsible for it. Someone who reads security advisories before they make the evening news, and who knows what to do when it matters.

That is the core of my WordPress maintenance: updates with good judgement, monitoring, and in an emergency somebody who reacts — before you even notice there was an emergency.

Frequently asked questions about hacked WordPress websites

Typical signs are unknown administrator accounts, PHP files in the uploads folder, warnings in Google Search Console, strange redirects and unusually slow load times. Many infections, however, stay deliberately invisible — if you suspect something, only an inspection of the files and server logs will tell.

No. An update closes the vulnerability, but it does not remove malware that was planted before the update. If a vulnerability has been actively exploited, the clean-up always includes a compromise assessment: comparing files against the originals, checking the user list, analysing the logs and renewing all credentials.

Only up to a point. A backup helps if it demonstrably dates from before the break-in and the vulnerability is closed immediately after restoring it — otherwise the restored website is reinfected within minutes. Since attackers often go unnoticed for weeks, backups are frequently already infected too.

That depends on the extent of the infection — from a few hours for an isolated case caught early to several days when backdoors are embedded on several levels. Far cheaper than any clean-up is prevention: regular maintenance with prompt updates and monitoring.

The most effective measures: prompt updates of core, plugins and themes; as few plugins as possible; strong, unique passwords with two-factor authentication; regular off-site backups — and a person who is responsible for following security advisories and reacts in an emergency.

Not sure whether your website is clean?

Drop me a line — a first assessment costs you nothing but five minutes.

Ask with no obligation