agentic

Introduction to Agentic

Agentic is a comprehensive standard library of artificial intelligence (AI) tools and functions. It's designed to seamlessly integrate with both standard TypeScript usage and applications that utilize large language models (LLMs). The versatility of Agentic allows it to work effortlessly with a variety of popular TypeScript AI software development kits (SDKs), including LangChain, LlamaIndex, Vercel AI SDK, and OpenAI SDK.

Key Features

One of the standout features of Agentic is its dual functionality. It can be employed as a regular TypeScript class or as an LLM-based tool where the LLM determines how and when to utilize these functions. This flexibility is achieved through the use of specific adapters that facilitate compatibility across different AI SDKs.

Here's a simple demonstration using the WeatherClient:

import { WeatherClient } from '@agentic/stdlib'

// Requires `process.env.WEATHER_API_KEY` (free from weatherapi.com)
const weather = new WeatherClient()

const result = await weather.getCurrentWeather({
  q: 'San Francisco'
})
console.log(result)

This example demonstrates how straightforward it is to retrieve weather data in San Francisco using the WeatherClient. Similarly, the library can be used in LLM-based applications, allowing more dynamic interaction where the choice of tool is determined by LLMs.

Integration with AI SDKs

Agentic provides excellent integration with various AI SDKs through specially designed adapters. This is especially beneficial for developers as it removes the hassle of creating additional code to bridge different systems. An example is provided below using Vercel's AI SDK:

// sdk-specific imports
import { openai } from '@ai-sdk/openai'
import { generateText } from 'ai'
import { createAISDKTools } from '@agentic/ai-sdk'

// sdk-agnostic imports
import { WeatherClient } from '@agentic/stdlib'

const weather = new WeatherClient()

const result = await generateText({
  model: openai('gpt-4o-mini'),
  tools: createAISDKTools(weather),
  toolChoice: 'required',
  prompt: 'What is the weather in San Francisco?'
})

console.log(result.toolResults[0])

By employing Agentic's adapters, developers can use this powerful AI standard library across a broad spectrum of projects without the need to develop any new integration layers.

Comprehensive Documentation

Full documentation for Agentic is available at agentic.so. This comprehensive resource provides everything from installation guides to detailed examples covering each AI SDK supported by Agentic.

Supported AI SDKs

Agentic supports a wide range of AI SDKs:

• Vercel AI SDK: Integration guides are readily available.
• LangChain: Detailed documentation ensures easy usage.
• LlamaIndex: Comprehensive support provided.
• Firebase Genkit: Seamlessly adaptable.
• Dexa Dexter: Supports advanced use cases.
• OpenAI: Easy integration with OpenAI models.

Each adapter-specific documentation is accessible through the agentic.so website.

Available Tools

Agentic comes with an array of tools and services, including but not limited to:

• Web Search: Tools like Bing and Jina offer enhanced search capabilities.
• Data Enrichment: Services such as Clearbit provide detailed information.
• Notification Systems: Novu facilitates diverse notification methods.

These tools provide versatile functionalities ranging from basic calculations to complex web scraping tasks, all accessible via Agentic’s interface.

Conclusion

Agentic positions itself as a robust AI standard library tailored for developers working with TypeScript and LLMs. Its ability to integrate across major AI SDKs makes it a valuable resource for building AI-driven applications seamlessly. Whether for weather data retrieval, news summarization, or web searches, Agentic simplifies and accelerates the development process, making it a compelling choice for developers in the AI space.