Setup external connectors for Knowledge Base

When you enabled the Knowledge Base Module for your TypingMind instance in the self-host environment, the only data source available is File Upload.

Here is how to enable external data source connectors like GitHub, Dropbox, Notion, etc.

ℹ️

Note: This setup step requires technical skills to edit the database. We will make it easier to setup within the Admin Panel in the future.

Updating configuration in the PostgreSQL database

Open your PostgreSQL database that you used for the Knowledge Base module, select the tbl_orgs table. There should be one record in that table, select it and update the props value of the that record with the following value:

json
{
  "availableSourceTypes": [
    "user_upload",
    "github",
    "notion",
    "web_scrape",
    "google_drive"
  ],
  "notionCredentials": {
    "clientID": "",
    "clientSecret": ""
  },
  "webScrapeCredentials": {
    "firecrawlAPIKey": ""
  },
  "googleDriveCredentials": {
    "apiKey": "",
    "clientID": "",
    "clientSecret": "",
    "projectNumber": ""
  }
}

In the property availableSourceTypes, identify the sources you want to enable using the values from the table below. Some data sources may require additional configurations as follow.

Data Source	JSON Value	Required config properties
Notion	`notion`	`notionCredentials`
GitHub	`github`	No other properties needed
Web Scrape	`web_scrape`	`webScrapeCredentials`
Google Drive	`google_drive`	`googleDriveCredentials`

Setup external data sources

Notion data source

Follow this guide to create a Notion integration.

During creating the integration, you will be asked to provide a Redirect URL, use the value below:

plain text
https://{DEPLOYMENT_HOST_NAME}/kb-api/api/v1/webhook/integrations/notion/callback

(Make sure to replace {DEPLOYMENT_HOST_NAME} with the actual host name of your self-host deployment.

Then add this new field to the value of props:

json
  "notionCredentials": {
    "clientID": "",
    "clientSecret": ""
  }

Web Scrape data source

Web scraping is done using an external service called Firecrawl. Go to https://www.firecrawl.dev/ to signup for an account and get an API key. Add this to the value of props:

json
  "webScrapeCredentials": {
    "firecrawlAPIKey": ""
  }

Google Drive data source

Create a new project and an API credentials in your Google Cloud account.

1. Create Google Cloud Project

Go to Google Cloud Console

Create a new project or select existing one

Note the Project Number (found in project settings) - this is required by the application

2. Enable Required APIs

Enable these APIs in your Google Cloud project:

Google Drive API - for file access and management

Google Picker API - for the file picker interface

3. Create OAuth2 Credentials

Go to APIs & Credentials > Credentials

Click "Create Credentials" > "OAuth 2.0 Client IDs"

Choose "Web application" as application type

Configure authorized redirect URIs:

plain text
https://{DEPLOYMENT_HOST_NAME}/kb-api/api/v1/webhook/integrations/google-drive/callback

4. Create API Key

Go to APIs & Credentials > Credentials

Click "Create Credentials" > "API Key"

Restrict the API key to:

Google Drive API

Google Picker API

Your domain/IP addresses

Required OAuth Scopes

The application uses this scope:

plain text
https://www.googleapis.com/auth/drive.file

Need help?

Feel free to contact support@typingmind.com to reach us.