A critical heap buffer overflow in Ruby's string case-conversion code, tracked as CVE-2026-132825 and dubbed ruby2shell, gives attackers a pre-authentication path to remote code execution in default Ruby on Rails and Sinatra application configurations1. The Ruby team disclosed the flaw on November 19, 2026, and shipped fixes in Ruby 3.3.14, 3.4.13, and 4.0.10.
Ruby rates the vulnerability Critical and says it is aware of proofs of concept that achieve remote code execution in some scenarios. Veria Labs says it will publish its own proof of concept roughly a week after disclosure.
Blast radius across self-hosted platforms
Veria Labs reports that self-hosted products bundling Ruby shipped an affected version in their latest release at the time of disclosure: GitLab 19.6.0, GitHub Enterprise Server 3.22.5, Mastodon 4.7.5, OpenProject 17.9.7, Discourse 2026.10.1, and Chatwoot 4.20.0.
Cloudflare says Ruby code that changes the case of strings from email, file imports, message queues, or background jobs can reach ruby2shell without an HTTP request, widening the attack surface beyond web traffic.
Cloudflare WAF response and its limits
Cloudflare added two rules to both the Cloudflare Free Managed Ruleset and the Cloudflare Managed Ruleset in the Cloudflare Web Application Firewall (WAF), with Block as the default action: one for ruby2shell and one for GHSA-87fm-7w84-j2px, a related integer overflow in the same string case-conversion code fixed in the same Ruby releases. The rules cover all customers, including those on the Free plan.
Cloudflare says the WAF is a compensating control, not a replacement for patching. Exploits can reach the bug through paths a WAF does not see, including requests sent directly to an origin IP address and traffic between internal services. Cloudflare says it has patched its own systems that use Ruby, found no indication of compromise, and continues to monitor attack traffic.
The same Ruby releases also fix six other memory-safety flaws, several of which Ruby says can lead to code execution.
ANALYSIS The pre-authentication, default-configuration attack surface across widely deployed platforms such as GitLab, GitHub Enterprise Server, and Mastodon compresses the patching window for operators of self-hosted instances.
Cloudflare's prior AI-infrastructure moves, including open-sourcing the Kitesurf browser engine for AI agents on November 13[2] and releasing open weights for extraction models on November 18[1], mean the company's own Ruby-dependent stack faced the same exposure it is now mitigating for customers.