BlahBlah Package

Introduction

Welcome to BlahBlah, an npm package that helps you automatically post comments on YouTube videos using your own YouTube account and OpenAI API key. This means full control - no third-party dependencies managing your credentials. You decide which account is used, ensuring transparency and security.

Overview

The package requires a valid configuration object with credentials for:

1. BlahBlah Project ID (from blahblah.dev)
2. OpenAI API Key
3. Google Client ID
4. Google Client Secret
5. Google Refresh Token

Once you have obtained these credentials, set them as environment variables (not hard-coded) and initialize the BlahBlah package in your Node.js project. The package will periodically (once-an-hour) search YouTube for newly published videos, and if certain criteria are met, it will post comments on those videos.

It is recommended to deploy the BlahBlah bot in a separate repository rather than integrating it into an existing project. The simplest approach is to clone this repository, deploy it, and configure your environment variables in your hosting platform. This ensures a clean setup and reduces potential conflicts with other application dependencies.

Installation

Install the BlahBlah package from npm:

npm install blah-blah

Or, if you're using Yarn:

yarn add blah-blah

Usage

Below is the most basic usage example:

import blahBlah from "blah-blah";

const BLAHBLAH_PROJECT_ID = process.env.BLAHBLAH_PROJECT_ID;
const OPENAI_API_KEY = process.env.OPENAI_API_KEY;
const GOOGLE_CLIENT_ID = process.env.GOOGLE_CLIENT_ID;
const GOOGLE_CLIENT_SECRET = process.env.GOOGLE_CLIENT_SECRET;
const GOOGLE_REFRESH_TOKEN = process.env.GOOGLE_REFRESH_TOKEN;

blahBlah({
  projectId: BLAHBLAH_PROJECT_ID,
  openAiApiKey: OPENAI_API_KEY,
  googleClientId: GOOGLE_CLIENT_ID,
  googleClientSecret: GOOGLE_CLIENT_SECRET,
  googleRefreshToken: GOOGLE_REFRESH_TOKEN,
  searchTerms: ["relevant", "keywords", "to-your", "product"],
});

Once this script is running, BlahBlah will:

1. Validate the provided configuration.
2. Periodically (once an hour) search for videos on YouTube that match your configured search terms.
3. Dynamically generate a comment using the OpenAI API and post it on each matched video.

Configuration & Credentials

This section will guide you through setting up each credential step by step.

1. BlahBlah Project ID

1. Visit blahblah.dev and create an account.
2. Create a new project.
3. Copy your Project ID.
4. Set it as an environment variable:

   BLAHBLAH_PROJECT_ID=your-project-id

2. OpenAI API Key

1. Go to OpenAI’s API Keys page (create an account if you don’t already have one), and make sure you top it up with some credits.
2. Generate a new secret key.
3. Copy your API key.
4. Set it as an environment variable:

   OPENAI_API_KEY=your-openai-api-key

3. Google Credentials

Before proceeding, decide which Google/YouTube account you want to use for this task. You may want to create a separate Google account specifically for this task to prevent any potential issues with your personal account (though there shouldn't be any, but it's safer).

To post comments via the YouTube Data API, follow these steps to set up a project in Google Cloud Console:

1. Create a Google Cloud Project (or select an existing project):

   • Go to Google Cloud Console
   • Click on Select a project (top-left) and choose New Project.
   • Enter a name, select a billing account if prompted, then click Create.

2. Enable the YouTube Data API:

   • In the Cloud Console, navigate to APIs & Services > Library.
   • Search for YouTube Data API v3.
   • Click Enable.

3. Create OAuth consent screen (if not already set up):

   • In APIs & Services, go to OAuth consent screen and click "Get Started".
   • Select External.
   • Add any required information (app name, support email, etc.) and save.

4. Create OAuth Client ID & Client Secret:

   • Go to Credentials.
   • Click Create Credentials, then OAuth client ID.
   • Select Web application.
   • Set the Redirect URI to:

     https://developers.google.com/oauthplayground

   • Click Create.
   • Copy the Client ID and Client Secret.
   • Set them as environment variables:

     GOOGLE_CLIENT_ID=your-google-client-id
     GOOGLE_CLIENT_SECRET=your-google-client-secret

5. Add your email as a test user

   • In the OAuth Consent Screen menu, navigate to "Audience".
   • Scroll down, and add your own email (yes yes, the one which owns this project) as a test user.

4. Generate the Refresh Token

To allow the script to post comments on your behalf (without manual user intervention every time), you need a Refresh Token.

1. Use the Google OAuth Playground:

   • Open Google OAuth Playground.
   • Click on the gear icon and check Use your own OAuth credentials.
   • Paste in your Client ID and Client Secret.
   • Use the following Scopes:

     https://www.googleapis.com/auth/youtube.force-ssl

   • Click Authorize APIs and allow the required permissions.

2. Exchange authorization code for tokens:

   • Click Exchange authorization code for tokens.
   • Copy the Refresh Token.
   • Set it as an environment variable:

     export GOOGLE_REFRESH_TOKEN="your-google-refresh-token"

Additional Configuration

In addition to the required properties, BlahBlah also supports optional configurations, such as:

• searchTerms: An array of words to be used in the YouTube search for fetching relevant videos (required).

• resultsPerJob: The maximum number of YouTube results fetched per search.

• publishTimeframe: An object containing min and max (in minutes), indicating how recent the videos should be:

  {
    publishTimeframe: {
      min: 30,
      max: 90
    }
  }

  In the configuration above, we will only be fetching videos which were posted between 30 and 90 minutes ago.

• llmConfig: If you wish to overwrite the AI instructions provided by BlahBlah, you can pass an object containing two fields:

  interface LLMConfig {
    modelInstructions: string;
    prompt: string;
  }

• openAiApiKey: By default, BlahBlah uses the model gpt-4o-mini-2024-07-18, but you may pass a different model in order to override this.

Troubleshooting

• Invalid Script Config: Ensure all required environment variables are set correctly.
• Comment not posting: Double-check that your Refresh Token is valid and that your credentials are correct. Also, ensure the YouTube Data API is enabled.
• API Errors: Check your usage and billing limits in both Google Cloud and OpenAI.