Proxy Helpers

When using the Proxy Gateway, you need to configure LLM SDK clients to route through Steward instead of calling APIs directly. These helpers simplify that process.

Manual Configuration

Most LLM SDKs support custom base URLs. Configure them to point at the Steward proxy:

OpenAI
Anthropic
Generic fetch

import OpenAI from "openai";

const openai = new OpenAI({
  apiKey: "steward",  // dummy — stripped by proxy
  baseURL: `${process.env.STEWARD_PROXY_URL}/openai/v1`,
});

// Use normally — proxy handles authentication
const completion = await openai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello" }],
});

import Anthropic from "@anthropic-ai/sdk";

const anthropic = new Anthropic({
  apiKey: "steward",
  baseURL: `${process.env.STEWARD_PROXY_URL}/anthropic`,
});

const message = await anthropic.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello" }],
});

async function proxiedFetch(
  path: string,
  init?: RequestInit
): Promise<Response> {
  const proxyUrl = process.env.STEWARD_PROXY_URL;
  const agentToken = process.env.STEWARD_AGENT_TOKEN;

  return fetch(`${proxyUrl}${path}`, {
    ...init,
    headers: {
      ...init?.headers,
      "Authorization": `Bearer ${agentToken}`,
      "Content-Type": "application/json",
    },
  });
}

// Usage
const response = await proxiedFetch("/openai/v1/chat/completions", {
  method: "POST",
  body: JSON.stringify({
    model: "gpt-4o",
    messages: [{ role: "user", content: "Hello" }],
  }),
});

StewardProxy Helper

The SDK provides a planned StewardProxy helper class for wrapping common SDK clients:

import { StewardProxy } from "@stwd/sdk";

const proxy = new StewardProxy({
  proxyUrl: process.env.STEWARD_PROXY_URL!,
  agentToken: process.env.STEWARD_AGENT_TOKEN!,
});

// Get pre-configured OpenAI client
const openai = proxy.openai();

// Get pre-configured Anthropic client
const anthropic = proxy.anthropic();

StewardProxy is planned for SDK v0.4.0. For now, use the manual configuration shown above — it’s straightforward and works with any SDK.

Environment Variables

In your agent container, you need:

STEWARD_PROXY_URL=http://steward-proxy:8080
STEWARD_AGENT_TOKEN=stwd_jwt_...

Pattern: Wrapper Function

A common pattern for projects with multiple API clients:

// lib/steward-proxy.ts
const PROXY_URL = process.env.STEWARD_PROXY_URL!;
const AGENT_TOKEN = process.env.STEWARD_AGENT_TOKEN!;

export function getOpenAI() {
  const OpenAI = require("openai").default;
  return new OpenAI({
    apiKey: "steward",
    baseURL: `${PROXY_URL}/openai/v1`,
    defaultHeaders: {
      "Authorization": `Bearer ${AGENT_TOKEN}`,
    },
  });
}

export function getAnthropic() {
  const Anthropic = require("@anthropic-ai/sdk").default;
  return new Anthropic({
    apiKey: "steward",
    baseURL: `${PROXY_URL}/anthropic`,
    defaultHeaders: {
      "Authorization": `Bearer ${AGENT_TOKEN}`,
    },
  });
}

// Anywhere in your agent code
import { getOpenAI, getAnthropic } from "./lib/steward-proxy";

const openai = getOpenAI();
const response = await openai.chat.completions.create({ ... });

Proxy Gateway Concepts — How the proxy works
Proxy API Reference — Proxy endpoints
Secrets Guide — Managing credentials

​Proxy Helpers

​Manual Configuration

​StewardProxy Helper

​Environment Variables

​Pattern: Wrapper Function

​Related

Proxy Helpers

Manual Configuration

StewardProxy Helper

Environment Variables

Pattern: Wrapper Function

Related