codex

Review code diffs against the base branch using the Codex CLI.

Updated Feb 28, 2026
One-click install
npx skills add https://github.com/Analog-Labs/pixelport-launchpad --skill codex-analog-labs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: codex
Source: https://github.com/Analog-Labs/pixelport-launchpad/tree/main/.agents/skills/gstack/codex
Command: npx skills add https://github.com/Analog-Labs/pixelport-launchpad --skill codex-analog-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides an independent, brutally honest second-opinion review of code changes against the base branch using the Codex CLI.

Core Features & Use Cases

  • Review mode: evaluate diffs against the base branch to validate changes.
  • Challenge mode: adversarial testing to uncover edge cases and failures.
  • Consult mode: quick, direct prompts for guidance or explanations.
  • Use cases include validating security, correctness, and robustness of changes across collaborative workflows.

Quick Start

Install Codex CLI, then run /codex review to generate an independent code review against your base branch.

Frequently Asked Questions about codex

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

FAQPage Schema
How do I get an independent second opinion on a code review using the Codex CLI?

To get a second opinion on a code review, run `/codex review` against your base branch. It evaluates diffs for security, correctness, and robustness, streaming the full output with a gate verdict and token count.

How does adversarial code testing work for finding edge cases in git diffs?

Adversarial code testing uses the Challenge mode to aggressively probe base branch diffs. It uncovers edge cases and potential failures by applying adversarial analysis to your code changes.

Do I need to install the Codex CLI to perform automated diff analysis?

Yes, installing the Codex CLI is required. Once installed, you can run commands to evaluate diffs, execute ad-hoc prompts, or perform plan-based reviews across your git repositories.

Can I use this for quick guidance on code changes without a full diff review?

Yes, you can use Consult mode for quick, direct prompts. It provides immediate guidance and explanations without needing to run a full base branch diff analysis.

Does this support cross-model comparisons for security-leaning code checks?

Yes, it supports cross-model comparisons during security-leaning code checks. The system persists session data and handles errors while streaming full output with a gate verdict.

What is the best way to validate the correctness of changes in collaborative workflows?

The best way to validate changes is using Review mode. It evaluates diffs against the base branch to independently validate security, correctness, and robustness across collaborative git workflows.