codex

Wrap the Codex CLI to review diffs and gate changes with verdicts.

Updated Feb 24, 2026
One-click install
npx skills add https://github.com/gaaschk/gaasch-family --skill codex-gaaschk
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex
Source: https://github.com/gaaschk/gaasch-family/tree/main/.claude/skills/gstack/codex
Command: npx skills add https://github.com/gaaschk/gaasch-family --skill codex-gaaschk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides an independent, brutally honest second opinion on code changes by wrapping the Codex CLI to review, challenge, or consult Codex across three modes.

Core Features & Use Cases

  • Code Review Mode: Compare the current diff against the base branch and gate the change with a pass/fail verdict.
  • Adversarial Challenge Mode: Prompt Codex to find edge cases, bugs, and security issues in the changes.
  • Consult Mode: Maintain a Codex conversation for follow-up questions with session continuity.
  • Use cases include pre-merge review, hardening code against edge cases, and obtaining an external perspective on architecture or prompts.

Quick Start

Ask Codex for a second opinion on your current diff by choosing review, challenge, or consult mode.

Frequently Asked Questions about codex

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How can I test edge cases and security issues in my code diff?

Adversarial challenge mode prompts Codex to find edge cases, bugs, and security issues in your changes. It actively probes your modifications to identify potential vulnerabilities and logic failures.

How do I maintain a session-based conversation for ongoing code questions?

Consult mode maintains a Codex conversation with session continuity for follow-up questions. It supports session-based discussions to obtain an external perspective on architecture or prompts.

What is required to run an adversarial code review using the Codex CLI?

Yes, you need Codex installed locally to use this wrapper. It operates by wrapping the Codex CLI to analyze diffs, test adversarial edge cases, and output structured gate verdicts with telemetry logging.

What does a structured code review output with gate verdicts and token estimates include?

Telemetry logging captures session data and token estimates during the code review process. The wrapper outputs structured results including gate verdicts and telemetry logs for tracking analysis.