@gentoro/sdk v0.0.8

Gentoro TypeScript SDK

Welcome to the Gentoro TypeScript SDK documentation. This guide will help you get started with integrating and using the SDK in your project.

Setup & Configuration

Supported Language Versions

This SDK is compatible with TypeScript >= 5.7.2.

Installation

To get started with the SDK, we recommend installing it using npm or yarn:

npm install @gentoro/sdk

or

yarn add @gentoro/sdk

Authentication

The Gentoro API uses an Access Token for authentication. This token must be provided to authenticate your requests to the API.

To obtain an API Key, register at Gentoro.

Setting the Access Token

When initializing the SDK, provide the configuration as follows:

const _config:SdkConfig = {
    apiKey: import.meta.env.VITE_GENTORO_API_KEY, // Your Gentoro API Key
    baseUrl: import.meta.env.VITE_GENTORO_BASE_URI, // Base Url where the Gentoro API is hosted
    authModBaseUrl: import.meta.env.VITE_GENTORO_AUTH_MOD_BASE_URI, // Base Url where the Gentoro Auth module UI is hosted
    provider: Providers.Openai, // The provider you want to use
};
const _gentoro:Gentoro = new Gentoro(_config);

Usage

Here's a basic example demonstrating how to authenticate and retrieve available tools:

import { Gentoro } from '@gentoro/sdk';
import * as dotenv from 'dotenv';

dotenv.config();

async function main() {
    // Initialize the Open AI client
    const _openai = new OpenAI.Client({
        apiKey: process.env['OPENAI_API_KEY'],
    });
    
    // Configure and Initialize Gentoro SDK
    const _config:SdkConfig = {
        apiKey: process.env['GENTORO_API_KEY'], // Your Gentoro API Key
        baseUrl: process.env['GENTORO_BASE_URI'], // Base Url where the Gentoro API is hosted
        authModBaseUrl: process.env['GENTORO_AUTH_MOD_BASE_URI'], // Base Url where the Gentoro Auth module UI is hosted
        provider: Providers.Openai, // The provider you want to use
    };
    const _gentoro:Gentoro = new Gentoro(_config);
    
    // Gather user inference content
    const _messages: ChatCompletionMessageParam[] = [{
        role: 'user',
        content: 'How many e-mails I currently have in my Gmail mailbox?',
    }];

    // Fetch tools from Gentoro.
    // Bridge UIDs are generated and captured at Gentoro's Studio.
    const _tools = await _gentoro.getTools(process.env['GENTORO_BRIDGE_UID'], _messages) as ChatCompletionTool[];

    // Invoke OpenAI API to generate completions, providing context (messages) and available tools
    let _completion:ChatCompletion  = await _client.chat.completions.create({
        messages: _messages,
        model: process.env['OPENAI_MODEL'], //gpt-4o-mini for example
        tools: _tools,
    });

    // Check if OpenAI requested tool_calls, if yes, delegates their execution to Gentoro, and capture the output
    _messages.push(... await _gentoro.runTools(_completion) as OpenAI.Chat.Completions.ChatCompletionMessageParam[])

    // Callback OpenAI with function call generated results
    _completion:ChatCompletion  = await _client.chat.completions.create({
        messages: _messages,
        model: process.env['OPENAI_MODEL'], //gpt-4o-mini for example
        tools: _tools,
    });

    // Print the completion    
    console.log(_completion.choices[0].message.content);
    // You have a total of 10 emails in your Gmail mailbox.
}

main();

Services

Methods

getTools(bridgeUid: string, messages: ChatCompletionMessageParam[]): Promise<ChatCompletionTool[]>

Fetches tools from a specific bridge (Bridge UIDs are captured from Gentoro's Studio).

const _messages: ChatCompletionMessageParam[] = ...;
const _tools = await _gentoro.getTools(process.env['GENTORO_BRIDGE_UID'], _messages) as ChatCompletionMessageParam[];

runTools(bridgeUid: string, completion: ChatCompletion): Promise

Analyzes the completion and runs the tools if any are requested. Returns pre-configured message with the tool_call message, and response.

const _completion:ChatCompletion  = await _client.chat.completions.create(...);
const _newMessages = await _gentoro.runTools(process.env['GENTORO_BRIDGE_UID'], _completion) as OpenAI.Chat.Completions.ChatCompletionMessageParam[];

Providers

A provider defines how SDK should handle and generate content.

export enum Providers {
    Openai,
    Anthropic,
    OpenaiAssistants,
    Vercel,
    Gentoro
}

License

This SDK is licensed under the Apache-2.0 License. See the LICENSE file for more details.