psf-optimizely-optimjack-check(8)

Script: /opt/psf-security-patches/optimizely/optimizely-optimjack-check.sh. Tool page: Optimizely OPTIMJACK supply-chain compromise. Source. With the package installed: man psf-optimizely-optimjack-check.

Optimizely Web Experimentation supply-chain compromise ("OPTIMJACK"): checker  (v1)
No CVE at time of writing (disclosed 16-17 Nov 2026 by Scott Helme / Report URI)

If this software is helpful, consider a donation to our 501c3 nonprofit:
https://privacysafe.foundation

On 16 November 2026 an attacker got write access to many unrelated customer
projects on Optimizely's Web Experimentation platform and injected a loader
into each project's "Project JavaScript" field, the one piece of custom code
that runs on every page of every experiment in that project. The file is
still served from cdn.optimizely.com with Optimizely's own certificate, so
nothing about where it comes from looks wrong: detection has to look at what
the file actually contains. Report URI found 188 infected snippets across
178 organisations (Zoom, the BBC, CNN, Salesforce, GitLab, IKEA, HSBC and
many others); Optimizely has since fixed the vulnerability and removed the
injected code, and by 16 Nov 23:11 UTC Report URI found every snippet clean.
The main command-and-control domain, 1688360.cc, was put on clientHold at
17 Nov 00:37 UTC and no longer resolves; this incident is winding down, not
live, but a site that was compromised still needs to check what was taken
and whether write access to its Optimizely account is still open.

What the injected code did: swapped cryptocurrency addresses in page text,
the clipboard and outgoing requests; skimmed payment card fields eMudhra's
hosted payment frames (Stripe, Adyen) were out of its reach, a site's own
form fields were not; sent cookies and page screenshots to the attacker;
and on iPhones and iPads loaded the Coruna (iOS 13-17) or DarkSword (iOS 18.x)
exploit kit through a hidden iframe. Three variants were seen: a plain
<script src> to the attacker's domain, the same URL hidden in base64 behind
atob() specifically so a text search for the domain would miss it, and a
full inline copy run through eval() as a fallback when the domain was
blocked. A restrictive script-src defeated the first and third; allowing
'unsafe-eval' or 'strict-dynamic' let either straight through, which is why
Report URI found only 24 of 256 sites they checked were fully protected.

"check" and "hunt" are read-only and make no changes. "check" reads one
Optimizely snippet (by URL or a local file you've already saved) and looks
for all three variants: the literal attacker domains, any atob() argument
that decodes to one of them, the known markers and telemetry channel codes
the payload set, and the two published SHA-256 builds of the payload file
itself. "hunt" does the same search across a directory of saved pages, JS
files or a CDN cache, for finding a copy you already have on disk. Nothing
here mitigates anything: there is no local fix, because the compromise was
on Optimizely's side and Optimizely has already removed the injected code.
What is left to do is site-side: see "What to do now" in the README.

Usage:
  ./optimizely-optimjack-check.sh [command] [options]

  check (default)  read-only. --url URL fetches and scans one snippet (needs
                   curl); --file PATH scans a file already saved to disk.
                   Exactly one of the two is required. Exit 0 = nothing
                   found, 1 = a known indicator found, 2 = could not be
                   fetched or read.
  hunt PATH ...     read-only. Scans each PATH (a file, or a directory
                   searched recursively) for the same indicators. Useful for
                   a saved CDN cache, a static export or a backup. Exit 0 =
                   nothing found, 1 = a lead, 2 = nothing readable at any PATH.
  csp              read-only. --header VALUE checks a Content-Security-Policy
                   header value you supply; --url URL fetches the header
                   from a live response instead. Reports whether it would
                   have blocked, partly blocked, or let through this
                   payload, the same four categories Report URI used.
  rules            prints a starting Content-Security-Policy line that would
                   have blocked every part of this payload.
  version          print the script version and exit

Env overrides:
  IOC_DOMAINS   space-separated, replaces the embedded attacker domain list
                (testing only; the built-in list is what this script checks
                by default and should not normally need replacing)

The Security Patches software and this documentation are dedicated to the public domain under CC0 1.0 Universal. Other pages on privacysafe.app/docs are CC BY-SA 4.0 unless the page says otherwise.

The PrivacySafe Ecosystem

Shop at Ivy Cyber

Ivy Cyber offers PrivacySafe products and services in the USA and has partnerships with 3NSoft in Canada and Kuip Limited in the European Union. Enterprise services and consulting are available at ivycyber.com

PrivacySafe Foundation is a 501(c)(3) nonprofit public charity that supports the development of the PrivacySafe applications, along with the components they rely upon: 3NWeb protocols, filesystem, documentation, and reference implementations.

PrivacySafe Foundation

Donations to PrivacySafe Foundation support education, research, documentation, and public code. Purchases from Ivy Cyber support PrivacySafe products, services, hosting, and technical support. Together, we're building a better web.