OpenAI launched Codex Security in November 2026, a suite of security tools including a command-line interface (CLI) and a TypeScript SDK designed to help developers automatically detect, validate, and fix security vulnerabilities in their codebases, enhancing software security. This information is based on the project's description.

Key Features and Workflow of Codex Security

Codex Security allows users to select security workflows on the Codex platform or use the CLI in local terminals to scan repositories, specific paths, or Git changes. Deep scans build upon standard scans and return deduplicated findings. The tool can validate candidate vulnerabilities, generate patches, and verify existing fixes.

Users can draft and update SECURITY.md policy files, save and browse threat models and scan results. The system assesses severity based on user rubrics and suggests owners from source and Git history. It also supports importing third-party scanner findings and publishing reports to the Codex Security Cloud platform.

The tool can track findings integrated with Linear, Jira, or GitHub, with export options for SARIF, JSON, and CSV formats, enabling automation in CI environments. Codex Security supports managing scans across multiple subprojects or multiple repositories centralized in campaigns.

Technical Requirements and Usage

Codex Security requires Node.js 22.13.0 or higher (supporting 22.x, 24.x, and 26.x) and Python 3.10 or later. Users install the CLI globally and sign in, or run via npx for a no-install option.

Users specify scan scope and time limits, for example, a 15-minute standard investigation allowance. The CLI provides commands to interpret reports with source evidence and exclusions.

The TypeScript SDK offers APIs for deep scan configuration, validation, severity classification, and result handling. Multiple linked git worktrees sharing the same Codex Security state directory share scan history and saved findings, improving cross-branch tracking efficiency.

Editor's take: Opportunities and Validation for Startups

OpenAI’s Codex Security introduces an automated and systematic approach to software security, helping startups reduce risks and maintenance costs while accelerating time to market. Its support for multiple formats and CI integration makes it suitable for startups managing multiple repositories and DevOps teams.

Entrepreneurs can validate the tool’s effectiveness by running initial scans in their development workflow, measuring accuracy in vulnerability detection and patch recommendations, and comparing with third-party scanners. Consideration of API cost and data privacy is essential prior to adoption since the tool relies on OpenAI models.