Anon Actions Library

Overview

This library showcases the power of Anon in building interactive AI agents that can perform actions on popular websites like LinkedIn and Amazon. Anon enables developers to access their users' sessions and execute actions on their behalf, opening up new possibilities for creating intelligent, automated applications.

What is Anon?

Anon is a groundbreaking platform that allows developers to build interactive AI agents by providing secure access to users' browser sessions. With Anon, you can:

• Create AI agents that interact with web applications on behalf of users
• Automate complex workflows across multiple websites
• Build innovative applications that leverage existing web platforms

This library demonstrates how Anon can be used to quickly implement common actions on popular websites, serving as a foundation for more complex AI-driven applications.

Features

This library provides ready-to-use browser scripts written in Playwright for common actions on consumer applications, such as LinkedIn and Amazon, showcasing how Anon can be used to build powerful applications on top of existing web platforms.

App Actions

Amazon

Instagram

Linkedin

Why Use This Library?

1. Rapid Development: These pre-built actions allow you to quickly implement complex web interactions in your AI agents or applications.
2. Real-World Examples: See how Anon can be used to interact with popular platforms, providing a blueprint for your own custom integrations.
3. Extensible Foundation: Use these scripts as a starting point to build more complex, AI-driven workflows and applications.

Prerequisites

• Anon SDK (including necessary credentials and setup)
• Node.js (version 18.0.0 or higher)
• Playwright

Installation

The Anon's actions library is currently in private beta. To get access and start building on Anon's platform:

1. Login to the Anon Console

2. Install the Anon actions library in your Node.js project:

   npm install @anon/actions \
   --registry=https://npm.cloudsmith.io/anon/anon-sdk/ \
   --//npm.cloudsmith.io/anon/anon-sdk/:_authToken=YQtlU86MhKOXijPS

You're now ready to use the Anon actions library!

See the "Usage" section below for examples on how to get started with your first Anon-powered application.

Usage

Here's how you can use these actions to quickly build interactive AI agents or automated workflows:

import { LinkedIn, NetworkHelper } from "@anon/actions";
import {
  Client,
  executeRuntimeScript,
  setupAnonBrowserWithContext,
} from "@anon/sdk-typescript";

const APP_USER_ID = process.env.ANON_APP_USER_ID!;
const API_KEY = process.env.ANON_API_KEY!;
const ENV = (process.env.ENV as "local" | "sandbox" | "prod" | "staging") || "sandbox";
const validEnvironments: Array<"local" | "sandbox" | "prod" | "staging"> = ["local", "sandbox", "prod", "staging"];
if (!validEnvironments.includes(ENV)) {
  throw new Error("Invalid env");
}

const account = {
  // check out our list of supported apps here: https://docs.anon.com/docs/getting-started/overview
  // this should align with a session you uploaded with the web-link example
  app: "linkedin",
  // this is the "sub" field of your user's JWTs
  userId: APP_USER_ID,
};

const client = new Client({
  environment: ENV,
  // create a server SdkClient and use its ApiKey
  // for testing, can alternately use an admin member's api key
  apiKey: API_KEY,
});

const main = async () => {
  console.log(
    `Setting up browser script for ${account.app} session for app user id ${APP_USER_ID}`,
  );
  const { browserContext } = await setupAnonBrowserWithContext(
    client,
    account,
    {
      type: "local",
      input: { headless: false, proxy: { isAnonProxyEnabled: true } },
    },
  );

  const networkHelper = new NetworkHelper(10000);

  const message = `This post was constructed with the Anon SDK! Integrate your AI apps to the real world using Anon. 
  Come check out https://www.anon.com/`
  await executeRuntimeScript({
    client,
    account,
    target: { browserContext },
    initialUrl: "https://www.linkedin.com",
    run: LinkedIn.createPost(networkHelper, message),
  });
};

main();

In this particular example, we are making a post on LinkedIn.

These examples demonstrate how you can quickly build interactive AI agents that perform actions on LinkedIn and Amazon using the Anon action library.

Building on Top of This Library

You can extend these actions to create more complex AI-driven applications. Some ideas:

1. A LinkedIn networking assistant that automatically engages with relevant connections
2. An Amazon price tracking and automatic purchasing agent
3. A cross-platform social media manager
4. An automated job application system using LinkedIn

If you have an idea on how to improve this library or want to contribute your own actions, please see the CONTRIBUTING.md guide.

Are you interested in learning more? Fill out this access form.