Agentsflare Codex Configuration Guide

Codex is OpenAI's official CLI coding tool that supports natural language instructions for code writing, debugging, and refactoring.

Installing Codex

Install globally via npm:

npm install -g @openai/codex

Or run directly with npx (no installation required):

npx @openai/codex

Verify installation:

codex --version

Configuring Codex

Codex reads API settings through environment variables or a configuration file. Below is how to configure it with the Agentsflare proxy.

Option 1: Environment Variables (Recommended)

Set the following environment variables in your terminal before starting Codex:

export OPENAI_API_KEY="Your Agentsflare API Key"
export OPENAI_BASE_URL="https://api.agentsflare.com/v1"

Then run Codex in the same terminal session:

codex

Option 2: Configuration File

Create or modify the Codex configuration file:

• macOS / Linux: ~/.codex/config.yaml
• Windows: %USERPROFILE%\.codex\config.yaml

Configuration file content:

model: gpt-5.2
api_key: "Your Agentsflare API Key"
base_url: "https://api.agentsflare.com/v1"

API Endpoint Reference

Agentsflare provides two OpenAI-compatible API endpoints. Choose based on your needs:

Endpoint	URL	Use Case
Chat Completions	https://api.agentsflare.com/v1/chat/completions	Standard chat, code generation, general tasks
Responses	https://api.agentsflare.com/v1/responses	Complex reasoning, tool calling, multi-turn context

Configure Chat Completions Endpoint

export OPENAI_API_KEY="Your Agentsflare API Key"
export OPENAI_BASE_URL="https://api.agentsflare.com/v1/chat/completions"

Configure Responses Endpoint

export OPENAI_API_KEY="Your Agentsflare API Key"
export OPENAI_BASE_URL="https://api.agentsflare.com/v1/responses"

Common Commands

Start an interactive session:

codex

Execute a single instruction directly:

codex "Refactor utils.js into TypeScript"

Launch with a specific model:

codex --model gpt-5.2

View help:

codex --help

Verify Configuration

After configuration, test the connection with:

codex "Hi, please briefly introduce yourself"

If Codex responds normally, the configuration is correct and you have successfully connected to OpenAI models through the Agentsflare proxy.