Documentation

CodeProxy User Guide

Learn how to use CodeProxy AI programming assistant to boost your development efficiency

By CodeProxy Editorial TeamPublished Updated

Agent guide (agent.md) Machine-readable index (llms.txt)

Codex

Standard integration docs and command examples for Codex CLI.

Quick Start

Choose your preferred installation method

Manual Installation

Install OpenAI Codex CLI tool using npm or brew

bash
npm install -g @openai/codex  # or use: brew install codex

Configuration

Codex requires creating a `.codex` folder in your system directory with required files.

~/.codex/config.toml

toml
model_provider = "CodeProxy"
model = "gpt-5.6-sol"
model_reasoning_effort = "medium"
disable_response_storage = true
model_context_window =1000000

[model_providers.CodeProxy]
name = "CodeProxy"
base_url = "https://codeproxy.dev/v1"
wire_api = "responses"
requires_openai_auth = true

~/.codex/auth.json

json
{
  "OPENAI_API_KEY": "your-api-key-here"
}

Quick API Key Setup

After preparing auth.json, jump to the dashboard `Config / keys` page to manage API keys.

Go to /keys

macOS/Linux: ~/.codex/

Windows: %USERPROFILE%\\.codex\\

Usage Examples

View help information

Show command help details

bash
codex help

Edit files

Edit a target file

bash
codex edit src/components/Button.tsx

Analyze code

Analyze a source directory

bash
codex analyze --files src/

Generate commit messages

Analyze code changes and generate commit messages automatically

bash
codex commit

Code refactoring

Refactor code with a specified design pattern

bash
codex refactor --pattern observer src/stores/

Resources

Visit NPM

Download the latest CodeProxy/Codex CLI package from npm

View Official Docs

Read complete official documentation for compatible tools