# openai/codex-security

Source: [GitHub](https://github.com/openai/codex-security)  
Feed7 permalink: https://feed7.dev/p/codex-security-0c0i99m  
Published: Unknown  
Trust: Needs Review (needs_review)

## Why Included

Codex Security adds CLI and TypeScript workflows for scanning code, validating findings, proposing fixes, and exporting results for CI or issue trackers.

## Source Summary

Codex Security provides a CLI and TypeScript SDK to scan repositories, selected paths, or Git changes. It can **validate findings, generate patches, and verify fixes**, and save reports with source evidence and exclusions. Results can be exported as SARIF, JSON, or CSV, or tracked through Linear, Jira, and GitHub integrations.

## Practical Implication

A builder can run scans locally or automate them in CI, then inspect the report's evidence and exclusions before asking an agent to patch a finding. The tooling also drafts `SECURITY.md` policies and stores threat models for later review.

## Agent-Ready Context

Codex Security provides a CLI and TypeScript SDK to scan repositories, selected paths, or Git changes. It can **validate findings, generate patches, and verify fixes**, and save reports with source evidence and exclusions. Results can be exported as SARIF, JSON, or CSV, or tracked through Linear, Jira, and GitHub integrations.

A builder can run scans locally or automate them in CI, then inspect the report's evidence and exclusions before asking an agent to patch a finding. The tooling also drafts `SECURITY.md` policies and stores threat models for later review.

The CLI requires Node.js 22.13.0+ within 22.x, or Node.js 24.x or 26.x, plus Python 3.10+. Some requests and protected findings require Trusted Access for Cyber; the local scan workflow does not require the optional Cloud connection.

## Connected Context

Feed7 judgment across 1059 accumulated Signals:

This packages security scanning with evidence-backed reports and a path from finding validation through patch generation to fix verification. It gives builders a local or CI workflow with export and tracking options, subject to specified runtime requirements and Trusted Access for some requests or findings.

- [alibaba/open-code-review](https://feed7.dev/p/open-code-review-0k2kdrb) — Both constrain agent review with structured evidence; Alibaba focuses on precise line-level comments and explicitly trades away recall, while Codex Security describes repository security scans, finding validation, and fix verification.
- [GPT-6.1 Sol helps CodeRabbit strengthen security across complex codebases](https://feed7.dev/p/gpt-6-1-sol-coderabbit-code-security-18cfhhh) — The CodeRabbit example establishes a security-review use case but leaves accuracy and fix validation unspecified; Codex Security documents validation and verification steps, though that alone does not establish their reliability.

## Context Map

- Layer: tools
- Domains: coding, security
- Topics: tool-use, coding-agents

## Uncertainty

- The CLI requires Node.js 22.13.0+ within 22.x, or Node.js 24.x or 26.x, plus Python 3.10+. Some requests and protected findings require Trusted Access for Cyber; the local scan workflow does not require the optional Cloud connection.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.
