codex-cli

Name: codex-cli
Availability: InStock
Author: tktaf

Execute local Codex CLI commands with a fixed GPT-5.4 model and high reasoning effort.

Updated Aug 23, 2026

Search Tags:#code review #cli #codex #long-running #reasoning #gpt-5.4

Ready to run

npx skills add https://github.com/tktaf/codex --skill codex-cli-tktaf

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: codex-cli
Source: https://github.com/tktaf/codex/tree/main/skills/codex-cli
Command: npx skills add https://github.com/tktaf/codex --skill codex-cli-tktaf

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill allows direct interaction with the local Codex CLI, bypassing the MCP, for deep coding and reasoning tasks that require a fixed model profile and high reasoning effort.

Core Features & Use Cases

Direct CLI Execution: Run Codex commands locally without relying on a middleware component.
Fixed Model Profile: Ensures consistent execution using gpt-5.4 with xhigh reasoning effort.
Long-Running Tasks: Suitable for extended operations like code reviews that need stable, uninterrupted processing.

Quick Start

Execute a long-running review prompt using the codex-cli skill with a 900-second timeout.

Frequently Asked Questions about codex-cli

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

FAQPage Schema

How do I run a long-running code review with high reasoning effort locally?▼

Direct CLI execution bypasses middleware components, sending commands straight to the local Codex CLI. It ensures consistent execution by fixing the model profile to GPT-5.4 with high reasoning effort for deep coding tasks.

Can I use Codex CLI for non-interactive deep coding tasks?▼

Yes, Codex CLI supports both interactive and non-interactive modes for deep coding and reasoning tasks. It maintains a fixed model profile and includes error handling for rate limits during execution.

What is the best way to handle rate limit errors during a long-running CLI code review?▼

The best way to handle rate limit errors is using the built-in error handling provided by the Codex CLI Skill. It manages rate limits natively while maintaining xhigh reasoning effort for uninterrupted code reviews.

Does the Codex CLI Skill require any middleware dependencies to function?▼

No, the Codex CLI Skill has zero dependencies and bypasses middleware components entirely. It allows direct interaction with the local Codex CLI to execute commands using a fixed GPT-5.4 model profile.

Why use a fixed model profile instead of a configurable one for local code reviews?▼

A fixed model profile ensures consistent execution using GPT-5.4 with xhigh reasoning effort. This stability is critical for long-running reviews that require deep coding and reasoning without model variation.