AI-found 'Node2Shell' flaw in Node.js Brotli decoder allows remote code execution
OpenAI Codex Security researchers found CVE-2026-111497, a critical out-of-bounds write in Node.js's bundled Brotli decoder that Imperva's Red Team dubbed 'Node2Shell' — a public exploit runs code on unpatched Node.js 24.22.0.
A critical remote code execution flaw in the Brotli decoder bundled with Node.js has a public exploit days after the fix shipped. Tracked as CVE-2026-111497 and dubbed “Node2Shell” by Imperva’s Red Team, it was found and reported by OpenAI Codex Security researchers Khai Tran and Sylvie Mayer, per the Node.js October 21 security advisory.
The bug is an out-of-bounds write (CWE-787, CWE-252): a crafted stream declaring a single block type then switching blocks was not rejected, so decoding continued in an invalid state and wrote past an internal buffer. Attacker-supplied Brotli data an app decompresses can corrupt process memory, causing a crash, memory disclosure or remote code execution.
Imperva’s proof of concept ran attacker code on the official Node.js 24.22.0 Linux arm64 image, The Hacker News reported on October 23. Google, the CVE numbering authority, scored it 9.2 Critical under CVSS v4.0; Node.js rates it High.
The fix is out: Node.js 22.23.4, 24.22.1 and 26.12.1 (October 21), plus Brotli library updates 1.0.10, 1.1.1 and 1.2.1 and Bun 1.4.4. Node.js 20 and older get no fix; the upstream fix was committed in February 2026 but shipped without a release or advisory. Only services that decompress untrusted Brotli data are reachable — Express 5 and body-parser 2 are routes into the vulnerable decoder, not the flaw’s source. As of October 23, no exploitation in the wild has been reported and the CVE is not in the CISA known-exploited catalogue, but with working exploit code public, unpatched servers should update now.
Read next: more on this story at our cybersecurity hub.
More on this topic: all Technology stories

