@glueco/plugin-llm-openai

OpenAI LLM plugin for Personal Resource Gateway. Provides typed access to OpenAI's GPT models including GPT-4o, GPT-4 Turbo, GPT-3.5 Turbo, and o1 reasoning models.

Features

• Full OpenAI API Support: Chat completions with all parameters (tools, streaming, response formats, etc.)
• Schema-First Validation: Request validation using Zod schemas with enforcement field extraction
• Typed Client: Full TypeScript support with autocomplete for target apps
• Dual-Entrypoint: Clean separation between proxy and client code

Installation

npm install @glueco/plugin-llm-openai

Usage

For Target Apps (Client-Side)

Use the typed client wrapper for full TypeScript support:

import { openai } from "@glueco/plugin-llm-openai/client";
import { GatewayClient, FileKeyStorage, FileConfigStorage } from "@glueco/sdk";

// Setup gateway client
const gatewayClient = new GatewayClient({
  keyStorage: new FileKeyStorage("./.gateway/keys.json"),
  configStorage: new FileConfigStorage("./.gateway/config.json"),
});

// Get transport and create typed client
const transport = await gatewayClient.getTransport();
const openaiClient = openai(transport);

// Chat completion (non-streaming)
const response = await openaiClient.chatCompletions({
  model: "gpt-4o",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "What is the capital of France?" },
  ],
  temperature: 0.7,
  max_tokens: 1000,
});

console.log(response.data.choices[0].message.content);

Streaming Chat Completion

const stream = await openaiClient.chatCompletionsStream({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Tell me a story" }],
});

const reader = stream.stream.getReader();
const decoder = new TextDecoder();

while (true) {
  const { done, value } = await reader.read();
  if (done) break;
  const chunk = decoder.decode(value);
  // Process SSE chunk
  console.log(chunk);
}

Tool Calling

const response = await openaiClient.chatCompletions({
  model: "gpt-4o",
  messages: [{ role: "user", content: "What's the weather in Paris?" }],
  tools: [
    {
      type: "function",
      function: {
        name: "get_weather",
        description: "Get current weather for a city",
        parameters: {
          type: "object",
          properties: {
            city: { type: "string" },
          },
          required: ["city"],
        },
      },
    },
  ],
});

const toolCall = response.data.choices[0].message.tool_calls?.[0];
if (toolCall) {
  console.log("Function called:", toolCall.function.name);
  console.log("Arguments:", JSON.parse(toolCall.function.arguments));
}

With OpenAI SDK (Compatibility Mode)

You can also use the standard OpenAI SDK with the gateway's fetch wrapper:

import OpenAI from "openai";
import { GatewayClient } from "@glueco/sdk";

const client = new GatewayClient({ ... });
const gatewayFetch = await client.getFetch();
const baseURL = await client.getResourceBaseUrl("llm", "openai");

const openai = new OpenAI({
  apiKey: "unused", // Gateway provides the key
  baseURL,
  fetch: gatewayFetch,
});

const response = await openai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello!" }],
});

Supported Models

GPT-4o Family

• gpt-4o - Latest GPT-4o
• gpt-4o-2024-11-20
• gpt-4o-2024-08-06
• gpt-4o-mini - Smaller, faster, cheaper

GPT-4 Turbo

• gpt-4-turbo - Latest GPT-4 Turbo
• gpt-4-turbo-2024-04-09
• gpt-4-turbo-preview

GPT-4

• gpt-4
• gpt-4-0613

GPT-3.5 Turbo

• gpt-3.5-turbo
• gpt-3.5-turbo-0125

o1 Reasoning Models

• o1 - Latest o1
• o1-preview
• o1-mini

Enforcement Fields

The plugin extracts these enforcement fields during validation:

| Field | Type | Description | | ----------------- | --------- | ------------------------------------------------ | | model | string | The requested model name | | stream | boolean | Whether streaming is requested | | usesTools | boolean | Whether tools/functions are used | | maxOutputTokens | number? | Max tokens (max_tokens or max_completion_tokens) |

These fields are used by the gateway's policy enforcement system to apply access controls.

Plugin Configuration

When setting up credentials in the gateway admin:

| Field | Type | Required | Description | | -------------- | ------ | -------- | --------------------------------------------- | | apiKey | secret | Yes | Your OpenAI API key (starts with sk-) | | organization | string | No | Optional organization ID | | baseUrl | url | No | Custom base URL (for Azure OpenAI or proxies) |

Actions

| Action | Description | | ------------------ | --------------------------------------- | | chat.completions | Create chat completions with GPT models |

License

MIT