Status: [LIVE]

Endpoints found:
https://api.openai.com/v1/chat/completions
https://api.openai.com/v1/chat

Source: https://github.com/xblanc33/CodexProxy

# Codex Proxy — OpenAI-Compatible Logging Proxy (Node/Express)

A tiny Node/Express proxy that forwards OpenAI‑compatible API requests while writing rich NDJSON logs for easy analysis. It supports standard JSON and streaming (SSE) responses and is handy for tracing prompts, parameters, and tool/function calls.

## Features

- OpenAI‑compatible proxy to any API base (default `https://api.openai.com`).
- Logs every request/response to `logs/requests.ndjson` (one JSON per line).
- Streams upstream SSE to the client while also logging the full response once.
- Extracts assistant content into a concise `content` field (best‑effort).
- Extracts tool/function calls into a `tool_calls` array for human‑readable auditing.
- Simple CORS for browser clients; health endpoint at `/health`.

## Requirements

- Node.js 18+ recommended. If you use Node <18, the proxy dynamically polyfills `fetch` via `node-fetch@3` (must be installed).

## Quick Start

1) Install dependencies

```bash
npm i express dotenv
# Optional on Node <18
npm i node-fetch@3
```

2) Create a `.env` file (example)

```ini
PORT=5000
API_BASE=https://api.openai.com/v1
#API_BASE=https://generativelanguage.googleapis.com
UPSTREAM_API_KEY=sk-...
ALLOW_CLIENT_AUTH=false
# Optional header helpers
OPENAI_ORG=
OPENAI_PROJECT=
OPENAI_BETA=
FORCE_OPENAI_HEADERS=false
# Logging
LOG_RAW_BODY=true
```

3) Run the proxy

```bash
node server.js
```

4) Verify it’s up

```bash
curl -sS http://localhost:5000/health
```

## Usage

Proxied endpoints (more can be added easily):
- `POST /v1/responses` (primary for Codex v0.139.0+)
- `POST /v1/chat/completions` (legacy OpenAI chat API)
- Catch‑all passthrough: `POST /v1/:first`

Examples:

```bash
# Non‑stream example
curl -sS -X POST \
  http://localhost:5000/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -H 'Authorization: Bearer $OPENAI_API_KEY' \
  -d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Hello"}]}'

# Stream example (SSE)
c