Creating a Slack chat bot

In this use case, you'll learn how to use serverless technologies to create a Slack bot that will run commands in a chat and respond to user messages.

Prepare the environmentPrepare the environment

1. Download the archive with the files required to create a bot.
2. If you do not have a folder yet, create one.
3. Create a service account and assign it the editor role for your folder.

Create an app and connect it to Yandex CloudCreate an app and connect it to Yandex Cloud

Register the Slack appRegister the Slack app

1. Log in to Slack. To create a bot, make sure you have workspace management permissions.
2. Create an app:
   1. Click Create a custom app.
   2. In the Name field, enter the app name: ServerlessBotApp.
   3. Select the available workspace and click Create app.
3. Grant permissions to the ServerlessBotApp app:
   1. In the app management menu, select OAuth & Permissions.
   2. Under Bot Token Scopes, add the following permissions: chat:write, commands, and im:history.
   3. After updating the permissions, reinstall the app: you'll see a notification at the top of the page. To reinstall the app, follow the link in the notification and click Allow.

Set up a link between Slack and Yandex CloudSet up a link between Slack and Yandex Cloud

1. Create an API gateway named for-slack-bot.

2. Create a function named for-slack-bot-challenge. Make sure it is private.

3. Create a function version:

   1. Create a file named index.py and paste the contents of the 0_for-slack-bot-challenge.py file from the archive into it.
   2. Specify the following parameters:
      • Runtime environment: python37
      • Entrypoint: index.handler
      • Timeout, sec: 5
      • Service account: Service account you created earlier

4. Edit the for-slack-bot API gateway. Add to the paths parameter the POST method configuration:

   paths:
     /:
       get:
         x-yc-apigateway-integration:
           type: dummy
           content:
             '*': Hello, World!
           http_code: 200
           http_headers:
             Content-Type: text/plain
       post:
         x-yc-apigateway-integration:
           type: cloud_functions
           function_id: <function ID>
           service_account_id: <service account ID>
         operationId: slack-challenge
   

   Parameters:

   • <Function ID>: ID of the for-slack-bot-challenge function.
   • <Service account ID>: ID of the service account.

Test the link between Slack and Yandex CloudTest the link between Slack and Yandex Cloud

1. Copy the service domain of the for-slack-bot API gateway.
2. Select the ServerlessBotApp app.
3. In the app management menu, select Event Subscriptions.
4. Set Enable Events to on.
5. In the Request URL field, insert the API gateway's address and wait for the Verified entry to appear.

Get a Token and Secret for the appGet a Token and Secret for the app

Select the ServerlessBotApp app:

• The Bot User OAuth Token value from the OAuth & Permissions section will be used for the SLACK_BOT_TOKEN environment variable.
• The Signing Secret value from the Basic Information section will be used for the SLACK_SIGNING_SECRET environment variable.

Create a databaseCreate a database

The bot will use responses to chat commands and messages taken from Yandex Managed Service for YDB. To enable this, prepare a table:

1. Create a database named for-slack-bot in Serverless mode.
2. In the for-slack-bot database, createa YDB table named coffee:

   1. Go to the Navigation tab.

   2. In the top-right corner, click New SQL query. The Query page opens.

   3. In the Query field, enter:

      CREATE TABLE coffee
      (
        id Utf8,
        name Utf8,
        PRIMARY KEY (id)
      );
      

   4. Click Run.

3. Add an entry to the table. For example, specify the coffee variety name and id = 1.

Create functionsCreate functions

Using functions, you can configure the bot's reactions to user actions in the chat. You'll create the following functions in this use case:

• For messaging between the bot and the user.
• For getting the bot's response to a simple command.
• For the bot to select a response to a command from Managed Service for YDB.

A function for messagingA function for messaging

1. Create a function named for-slack-bot-small-talk. Make sure it is private.

2. Create a function version:

   1. Create a file named requirements.txt and specify the following libraries in it:

      slack_sdk
      slack_bolt
      boto3
      

   2. Create a file named index.py and paste the contents of the 1_for-slack-bot-small-talk.py file from the archive into it.

   3. Specify the following parameters:

      • Runtime environment: python37
      • Entrypoint: index.handler
      • Timeout, sec: 5
      • Service account: Service account you created earlier

   4. Add environment variables:

      • SLACK_BOT_TOKEN
      • SLACK_SIGNING_SECRET

A function for responding to commandsA function for responding to commands

1. Create a function named for-slack-bot-hello-from-serverless. Make sure it is private.

2. Create a function version:

   1. Create a file named requirements.txt and specify the following libraries in it:

      slack_sdk
      slack_bolt
      boto3
      

   2. Create a file named index.py and paste the contents of the 2_for-slack-bot-hello-from-serverless.py file from the archive into it.

   3. Specify the following parameters:

      • Runtime environment: python37
      • Entrypoint: index.handler
      • Timeout, sec: 5
      • Service account: Service account you created earlier

   4. Add environment variables:

      • SLACK_BOT_TOKEN
      • SLACK_SIGNING_SECRET

A function for selecting responses to commandsA function for selecting responses to commands

1. Create a function named for-slack-bot-what-kind-of-coffee. Make sure it is private.

2. Create a function version:

   1. Create a file named requirements.txt and specify the libraries and the version of Managed Service for YDB in it:

      slack_sdk
      slack_bolt
      boto3
      ydb==0.0.41
      

   2. Create a file named index.py and paste the contents of the 3_for-slack-bot-what-kind-of-coffee.py file from the archive into it.

   3. Specify the following parameters:

      • Runtime environment: python37
      • Entrypoint: index.handler
      • Timeout, sec: 5
      • Service account: Service account you created earlier

   4. Add environment variables:

      • SLACK_BOT_TOKEN
      • SLACK_SIGNING_SECRET

   5. Add the variables to work with Managed Service for YDB:

      • ENDPOINT: First part of the Endpoint field value from the Overview section in the for-slack-bot database properties (the one preceding /?database=), e.g., grpcs://ydb.serverless.yandexcloud.net:2135.
      • DATABASE: Second part of the Endpoint field value from the Overview section in the for-slack-bot database properties (the one following /?database=), e.g., /ru-central1/r1gra875baommfd5leds/g5n22e7ejfr16h9oif9d.
      • USE_METADATA_CREDENTIALS = 1.

Edit the API gatewayEdit the API gateway

To make sure the bot starts responding to user messages, link the created functions to the app. To do this, edit the for-slack-bot API gateway and add to the paths parameter the POST method configurations:

paths:
  /:
    post:
      x-yc-apigateway-integration:
        type: cloud_functions
        function_id: <ID_of_function_1>
        service_account_id: <Service account ID>
      operationId: slack-challenge
  /hello-from-serverless:
    post:
      x-yc-apigateway-integration:
        type: cloud_functions
        function_id: <ID_of_function_2>
         service_account_id: <Service account ID>
       operationId: hello-from-serverless
  /what-kind-of-coffee:
    post:
      x-yc-apigateway-integration:
        type: cloud_functions
        function_id: <ID_of_function_3>
        service_account_id: <Service account ID>
      operationId: /what-kind-of-coffee

Parameters:

• <Service account ID>: ID of the service account.
• <ID_of_function_1>: ID of the for-slack-bot-small-talk function.
• <ID_of_function_2>: ID of the for-slack-bot-hello-from-serverless function.
• <ID_of_function_3>: ID of the for-slack-bot-what-kind-of-coffee function.

Add commands to SlackAdd commands to Slack

Using commands, you can set up the bot's actions in the chat. Commands start with a / character and their list is always visible to the user.

1. Select the ServerlessBotApp app.
2. In the app management menu, select Slash Commands and click Create New Command.
3. Add a command for the for-slack-bot-hello-from-serverless function:
   • In the Command field, enter /hello-from-serverless.
   • In the Request URL field, paste the url from the for-slack-bot API gateway specification, adding the /hello-from-serverless command's URL to it.
   • In the Short descriptions field, enter any short description of the command.
   • Click Save.
4. Add a command for the for-slack-bot-what-kind-of-coffee function:
   • In the Command field, enter /what-kind-of-coffee.
   • In the Request URL field, paste the url from the for-slack-bot API gateway specification, adding the /what-kind-of-coffee command's URL to it.
   • In the Short descriptions field, enter any short description of the command.
   • Click Save.
5. After adding the new commands, reinstall the app: you'll see a notification at the top of the page. To reinstall the app, follow the link in the notification and click Allow.

Test the Slack botTest the Slack bot

Open the Slack client and choose a chat with the ServerlessBotApp bot under Apps.

1. To test the for-slack-bot-small-talk function:
   • Send a message saying :wave: in the chat. The bot should respond Hi there, @<username>!.
   • Send a message saying knock knock in the chat. The bot should respond Who's there?.
2. To test the for-slack-bot-hello-from-serverless function:
   • Send the /hello-from-serverless command in the chat. The bot should respond Thanks!.
3. To test the for-slack-bot-what-kind-of-coffee function:
   • Send the /what-kind-of-coffee command in the chat. The bot should respond Today we use <entry from the coffee table>.