← AI agent vulnerabilities

CVE-2026-19591

Codex

8.8
Severity
High
Published
September 1, 2026
Last modified
September 2, 2026

OpenAI Codex CLI for Windows, macOS, and Linux and Codex Desktop for Windows and macOS misclassified certain PowerShell commands as safe because their command-safety parser interpreted PowerShell's stop-parsing token (--%) differently than PowerShell itself. If a user opens an attacker-prepared repository and Codex follows its instructions, Codex can run a file-writing Git command without requesting user approval. On macOS and Linux, exploitation additionally requires separately installed PowerShell Core (pwsh) to be invoked. If filesystem protections permit the write, the command can modify Codex's configuration. If Codex later loads the modified configuration, it can launch an attacker-controlled MCP server and execute code with the user's privileges, allowing it to read, change, or delete files accessible to that account. The approval bypass does not disable filesystem sandboxing; the default filesystem sandbox on macOS and Linux can prevent writes outside permitted locations.

CVSS vector CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

Weakness

References

CVE-2026-19591 on NVD

Listed in

High

More Codex CVEs

Sep 1
CVE-2026-195927.3

OpenAI Codex CLI for Windows, macOS, and Linux and Codex Desktop for Windows and macOS automatically collected Git repository metadata without disabling the repository-local core.fsmonitor setting.

Apr 14
CVE-2025-612609.8

A vulnerability was identified in OpenAI Codex CLI v0.23.0 and before that enables code execution through malicious MCP (Model Context Protocol) configuration files.

Sep 22
CVE-2025-595328.6

Due to a bug in the sandbox configuration logic, Codex CLI could treat a model-generated cwd as the sandbox’s writable root, including paths outside of the folder where the user started their session.

Aug 13
CVE-2025-553458.8

Using Codex CLI in workspace-write mode inside a malicious context (repo, directory, etc) could lead to arbitrary file overwrite and potentially remote code execution due to symlinks being followed outside…

Jul 25
CVE-2025-545584.1

OpenAI Codex CLI before 0.9.0 auto-approves ripgrep (aka rg) execution even with the --pre or --hostname-bin or --search-zip or -z flag.