Does this skill support specific git branches?

Yes, it can compare your current code against a base branch (like main) to perform targeted code reviews on new logic.

What is the difference between simple and complex tasks in this skill?

Simple tasks involve local modifications and standard refactoring, while complex tasks involve system-level design, multi-constraint problems, and deep performance analysis.

How do I trigger a high-intensity reasoning session?

You can trigger it by mentioning tasks like 'architecture analysis' or 'performance optimization', which prompts the use of gpt-5.2 with the xhigh reasoning effort parameter.

Can I use Codex-AI to review code before I commit it?

Yes, the skill includes a specific workflow for reviewing uncommitted changes using the --uncommitted flag within the Codex CLI.

What models does the Codex-AI skill use for code analysis?

It primarily utilizes high-reasoning models such as gpt-5.1-codex-max for general reviews and gpt-5.2 for complex architectural and performance tasks.

Codex-AI Advanced Code Intelligence

Name: Codex-AI Advanced Code Intelligence
Author: junjiangao

byjunjiangao

Collaboration Tools

Automates deep code reviews, algorithm design, and architectural analysis using high-reasoning Codex models.

About

Codex-AI is a specialized skill for Claude Code that empowers developers to perform high-level technical tasks such as code reviews, algorithm design, and system architecture analysis through the Codex CLI. By integrating advanced models like gpt-5.1-codex-max and gpt-5.2 with adjustable reasoning intensity, it provides a sophisticated environment for solving complex engineering problems and identifying performance bottlenecks. This skill is particularly useful for senior developers and architects who require deep, context-aware insights into their codebase that go beyond basic linting or simple logic suggestions.

Key Features

Deep reasoning effort configuration to solve multi-constraint technical challenges.
Architectural evaluation capabilities for system-level design and scalability audits.
Targeted performance optimization tools to detect and resolve logic bottlenecks.
Multi-model support including gpt-5.1-codex-max and gpt-5.2 for varied task complexity.
0 GitHub stars
Comprehensive code review workflows for pull requests, specific commits, and local changes.

Use Cases

Analyzing existing project architecture to identify performance bottlenecks and security vulnerabilities.
Performing rigorous, high-reasoning code reviews on uncommitted local changes before a PR.
Designing complex backend algorithms that require high-intensity logic and system-level constraints.

About

Key Features

Deep reasoning effort configuration to solve multi-constraint technical challenges.
Architectural evaluation capabilities for system-level design and scalability audits.
Targeted performance optimization tools to detect and resolve logic bottlenecks.
Multi-model support including gpt-5.1-codex-max and gpt-5.2 for varied task complexity.
0 GitHub stars
Comprehensive code review workflows for pull requests, specific commits, and local changes.

Use Cases

Analyzing existing project architecture to identify performance bottlenecks and security vulnerabilities.
Performing rigorous, high-reasoning code reviews on uncommitted local changes before a PR.
Designing complex backend algorithms that require high-intensity logic and system-level constraints.