CVE-2026-19590: OpenAI Codex Desktop vulnerability
OpenAI Codex Desktop for Windows and macOS could execute attacker-controlled Git hooks because automated Git operations trusted the repository's local core.hooksPath setting. If a user opens an attacker-prepared repository whose preserved .git/config points core.hooksPath to an attacker-controlled directory, Codex can run a malicious hook while processing the repository. The hook executes outside Codex's command sandbox, without user approval, and with the user's privileges, allowing it to read, change, or delete the user's files and access other resources available to the user's account. An ordinary Git clone does not preserve the attacker-controlled repository-local configuration required for exploitation.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
OpenAI Codex Desktop (macOS)to a version that resolves this vulnerability.Fixed in 26.519.22136 - Upgrade
Upgrade
OpenAI Codex Desktop (Windows)to a version that resolves this vulnerability.Fixed in 26.519.21041Patch Microsoft Store package 26.519.2081.0
Event History
Frequently Asked Questions
Who is realistically exposed to this issue?
Users of Codex Desktop on Windows or macOS are exposed when they open an attacker-prepared repository whose existing .git/config preserves a repository-local core.hooksPath pointing to an attacker-controlled directory. A repository obtained through an ordinary Git clone does not retain the required configuration.
What must an attacker provide for exploitation to occur?
The attacker needs to provide a repository with a preserved local Git configuration that sets core.hooksPath to an attacker-controlled directory, along with a malicious hook in that location. Codex must then perform an automated Git operation while processing that repository.
Does exploitation require user approval or execution inside Codex's command sandbox?
No. The malicious hook can run without user approval and executes outside Codex's command sandbox, with the privileges of the affected user's account.
What could a successful hook execution access or modify?
The hook can read, modify, or delete the user's files and access other resources available to that user's account.