codex

Name: codex
Availability: InStock
Author: txema-puch

Review code diffs with OpenAI Codex CLI pass/fail gates.

1|Updated Mar 23, 2026

Search Tags:#security #cli #codex #bash #review #ai-assistant #code-review

Ready to run

npx skills add https://github.com/txema-puch/drone-ai-saturdays --skill codex-txema-puch

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: codex
Source: https://github.com/txema-puch/drone-ai-saturdays/tree/main/.claude/skills/gstack/codex
Command: npx skills add https://github.com/txema-puch/drone-ai-saturdays --skill codex-txema-puch

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The Codex-based wrapper provides an independent, blunt assessment of code changes, surfacing issues that might be missed by standard reviews and enabling more robust gate decisions.

Core Features & Use Cases

Code review mode with pass/fail gates to evaluate diffs against the base branch.
Adversarial challenge mode that probes edge cases, performance pitfalls, and potential vulnerabilities.
Consult mode for interactive Q&A with Codex, preserving session continuity for follow-ups.

Quick Start

Run the codex tool with a chosen mode (review, challenge, or consult) to process your current diff against the base branch and read the gate result.

Frequently Asked Questions about codex

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

FAQPage Schema

How do I get an automated second opinion on my code changes before merging?▼

To get an automated second opinion on code changes, you can use the Codex CLI wrapper to evaluate your current diff against the base branch. It provides a brutally honest review with pass or fail gate decisions to improve robustness.

How can I use an AI assistant to probe edge cases and security vulnerabilities in my code?▼

You can use an AI assistant to probe edge cases and security vulnerabilities by running the adversarial challenge mode. This mode probes performance pitfalls and potential vulnerabilities to surface issues missed by standard reviews.

Do I need the OpenAI Codex CLI installed to run automated code reviews on my branches?▼

Yes, you need the OpenAI Codex CLI installed and available in your environment. The wrapper requires Codex CLI availability to process diffs and generate telemetry for usage logging and gate decisions.

Can I ask follow-up questions during an AI code review without losing session context?▼

Yes, you can ask follow-up questions without losing session context by using the consult mode. This mode enables interactive Q&A with Codex while preserving session continuity for deeper testing.

What is the best way to automate a pass or fail gate decision for a code diff?▼

The best way to automate a pass or fail gate decision for a code diff is using the review mode. It evaluates your current diff against the base branch and outputs a gate result to block or approve changes.