CVE-2026-126348: Addressable punycode decoding host spoofing

CVE-2026-126348 · Severity: medium · CVSS 4.3 · Published 2026-11-12

Addressable is a Ruby library for parsing and manipulating URLs. The library's punycode decoder (used to convert internationalized domain names) incorrectly processes invalid domain labels, potentially allowing an attacker to craft a URL that displays a trusted hostname while actually connecting to an attacker-controlled server. This could enable phishing, man-in-the-middle attacks, or bypassing host-based access controls.

Technologies: Addressable, addressable (RubyGems). Vendors: Addressable, RubyGems.

Quick answers

Is CVE-2026-126348 exploited in the wild?
No. Junglewise Threat Intelligence has no report of CVE-2026-126348 being exploited in the wild.
How severe is CVE-2026-126348?
Medium severity, with a CVSS score of 4.3, published 12 November 2026.
Which products and versions are affected?
Addressable 2.0.0 to 2.9.0; addressable (RubyGems) >= 2.0.0, < 2.9.1.
Is there a fix for CVE-2026-126348?
Yes: fixed in 2.9.1. Junglewise finds every place you run it, drafts the upgrade and applies it once you approve.
How is CVE-2026-126348 attacked?
It is attacked over the network.
Am I affected by CVE-2026-126348?
Check whether you run Addressable or addressable (RubyGems) at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

How to fix

  • Fixed in: 2.9.1

Technical details

The pure-Ruby implementation of Addressable::IDNA.to_unicode decodes invalid Punycode A-labels instead of rejecting them, causing labels that decode to ASCII text or fail round-trip conversion to be displayed differently from their actual connection target. An attacker can craft Punycode labels (e.g., xn--google-) that decode to trusted-looking ASCII hostnames while retaining the encoded attacker-controlled host for the actual connection. The native libidn backend and other methods are unaffected; the vulnerability requires code to rely on the pure-Ruby backend and trust display_uri output for host validation.

Affected products

  • Addressable Addressable 2.0.0 to 2.9.0

Identifiers

  • CVE-2026-126348
  • GHSA-9crm-g4mf-wc5j

Timeline

  • 2026-11-12: advisory

References

This page as Markdown. Free to reuse under CC BY 4.0 with a link back.