OpenAI Codex security directory

Codex Security

Published by OpenAI

High risk

Provides workflows for reviewing repositories and code changes for security issues. Model threats, validate findings and assess patches, then remediate or verify fixes. Create vulnerability reports and track findings in development tools.

Security assessment

Plutonium assessed this plugin for permissions, capabilities, and security-relevant behavior.

  • Edits source to fix findings and writes SECURITY.md: Can read, create, or change files on the local filesystem or in a selected project.
  • Builds targets and runs PoCs, debuggers and repo tests: Can execute code, commands, scripts, unrestricted queries, or dynamic gateway actions.
  • Creates GitHub issues, Jira/Linear issues and draft security advisories as the user: Can send messages, comments, invitations, posts, or published content attributed to the connected user.

Available capabilities

This add-on exposes 15 tools or capabilities.

  • assess-patch-risk
  • attack-path-analysis
  • deep-security-scan
  • define-security-policy
  • finding-discovery
  • fix-finding
  • propose-security-hardening
  • security-diff-scan
  • security-scan
  • threat-model
  • track-findings
  • triage-finding
  • validation
  • verify-fix
  • vulnerability-writeup

The interactive security report will load automatically.