skills.rest
VertaKhan
VertaKhanCOMMUNITY
Explore all skills by this author →
GitHub

codex

Runs OpenAI Codex CLI for independent code review, adversarial challenge, and consult sessions.

Updated Apr 5, 2026
Search Tags:#code review#OpenAI#cli#codex#git diff#second opinion#adversarial review
One-click install
npx skills add https://github.com/VertaKhan/cs_gstack --skill codex-vertakhan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: codex
Source: https://github.com/VertaKhan/cs_gstack/tree/main/.claude/skills/gstack/codex
Command: npx skills add https://github.com/VertaKhan/cs_gstack --skill codex-vertakhan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @openai/codex.

What problem does it solve? Getting a second opinion on code changes from the same AI that wrote them creates blind spots. This Skill wraps the OpenAI Codex CLI to provide an independent, cross-model review of your diffs, an adversarial challenge mode that tries to break your code, and a consult mode for open-ended technical questions with session continuity. ## Core Features & Use Cases - Review Mode: Runs codex review against the current branch diff with a pass/fail gate based on critical [P1] findings, plus cross-model comparison when Claude's own /review was already run. - Challenge Mode: Adversarially attacks your diff to find edge cases, race conditions, and failure modes before they hit production. - Consult Mode: Asks Codex anything with session continuity for follow-up questions, using medium reasoning effort for interactive speed. - Use Case: Before merging a PR, run /codex review to get an independent verdict with a GATE: PASS/FAIL result, token cost, and a persisted review log entry that feeds into the plan file's review report. ## Quick Start Ask the AI to run /codex review on the current branch to get an independent second opinion on your diff with a pass/fail gate.

Frequently Asked Questions about codex

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

FAQPage Schema
How do I get a second AI opinion on a code review?▼

Run /codex review to send your current branch diff to the OpenAI Codex CLI for independent review. The output includes a GATE: PASS or FAIL verdict based on whether any critical [P1] findings are detected, plus token usage and estimated cost.

How to adversarially test my code changes with AI?▼

Use /codex challenge with an optional focus area. Challenge mode runs Codex in adversarial mode against your diff, trying to break your code by finding edge cases and failure modes rather than just reviewing style.

Does codex review work without the Codex CLI installed?▼

No, the skill checks for the codex binary first and stops if it is not found. Install it with npm install -g @openai/codex, then rerun the command.

What is the difference between codex review and challenge mode?▼

Review mode produces a code review with a pass/fail gate based on [P1] critical findings. Challenge mode is adversarial: it actively tries to break your code. Both default to high reasoning effort, while consult mode uses medium for faster interactive responses.

Can I increase Codex reasoning effort for tougher reviews?▼

Yes, pass --xhigh anywhere in your command input. The skill strips the flag from the prompt and sets model_reasoning_effort to xhigh for all modes, overriding the per-mode defaults of high for review/challenge and medium for consult.

Related Skills

Explore More Software Engineering Skills→
bastani-incbastani-inc
484

gh-create-pr

Commit unstaged changes, push the branch, and create a pull request.

Official
Advanced
badbsallyybadbsallyy

prepare-pr

Rebase a GitHub pull request onto main and run automated checks.

Community
Advanced
Happyverse-TeamHappyverse-Team

find-code-tasks

Aggregate frontmatter from .code-task.md files and export filtered task lists.

Official
Intermediate
eklyukineklyukin

codex

Obtains independent second opinions from OpenAI Codex CLI for planning, code review, and implementation review.

Community
Intermediate