CLI Agent OpenAI Adapter — AI skill for Claude Code

cli-agent-openai-adapter

Convert CLI-based AI agents (Claude Code, etc.) to OpenAI ChatAPI endpoints.

Overview

This adapter allows you to use local CLI tools like Claude Code as drop-in replacements for OpenAI's API in your development environment, while keeping the same code structure for production.

**Use Cases:**

• Production: Use OpenAI API (pay per token)
• Development: Use local Claude Code with Haiku model (reduce costs)
• Same Code: Switch between environments using the same API interface (e.g., LangChain's ChatOpenAI)

**Default Model:** This adapter uses Claude Haiku by default for cost efficiency during development. You can configure a different model (e.g., Sonnet, Opus) via the `MODEL` environment variable.

Features

• ✅ OpenAI-compatible API endpoints (/v1/chat/completions)
• ✅ Support for conversation history
• ✅ Stateless execution (like OpenAI API)
• ✅ Chat-only mode (tools disabled for safety)
• ✅ TypeScript with full type definitions
• 🚧 Claude Code adapter (initial implementation)
• 🔜 Codex adapter (future)
• 🔜 Gemini CLI adapter (future)

Demo

Try the adapter with the minimal, dependency-free web client:

# Start the adapter (project root)
npm ci
npm run build && npm start

# Open the client in your browser
# File path: examples/minimal-web-client/index.html

[→ Minimal Web Client README](examples/minimal-web-client/README.md)

Installation

npm install -g cli-agent-openai-adapter

Or use directly with npx:

npx cli-agent-openai-adapter

Prerequisites

• Node.js >= 20.0.0
• Claude Code CLI installed and accessible in PATH

To verify Claude Code is installed:

claude --version

Usage

Start the Server

cli-agent-openai-adapter

By default, the server starts at `http://localhost:8000`.

Configuration

Configure using environment variables:

export ADAPTER_TYPE=claude-code  # Adapter to use
export MODEL=haiku                # Cla