How to translate Slack messages with Power Automate

My flows > + New flow > Automated cloud flow

Create new automated flow

Go to make.powerautomate.com and sign in with your work account. Click 'My flows' in the left sidebar, then the blue '+ New flow' button. Select 'Automated cloud flow' from the dropdown menu. Name it 'Slack Message Translator' and search for 'When a reaction is added to a message' in the trigger list.

1. 1Click '+ New flow' in the top navigation
2. 2Select 'Automated cloud flow'
3. 3Type 'Slack Message Translator' as the flow name
4. 4Search for 'reaction' in the trigger search box
5. 5Select 'When a reaction is added to a message (Slack)'

Slack trigger > Sign in

Connect your Slack workspace

Click 'Sign in' on the Slack trigger card. You'll be redirected to Slack's OAuth screen asking for permissions. Make sure you're signed into the correct Slack workspace before approving. Grant permissions for reading messages, reactions, and posting replies. Power Automate needs these scopes to detect flag reactions and post translations back to channels.

1. 1Click the 'Sign in' button on the Slack connector
2. 2Select your Slack workspace from the list
3. 3Review the requested permissions carefully
4. 4Click 'Allow' to authorize Power Automate
5. 5Wait for the green 'Connected' status

Slack trigger > Configure parameters

Configure reaction trigger settings

Leave the 'Team' field as your connected workspace. For 'Channel', click the dropdown and select 'Any channel' to monitor all channels, or pick specific channels if you only want translation in certain places. The 'User' field should remain empty to catch reactions from anyone. This setup makes the flow trigger whenever someone adds any emoji reaction to any message.

1. 1Set Team to your connected workspace
2. 2Click the Channel dropdown
3. 3Select 'Any channel' for workspace-wide coverage
4. 4Leave User field blank
5. 5Leave Reaction field empty to catch all emoji types

New step > Control > Condition

Add condition to filter flag emojis

Click '+ New step' below the trigger and search for 'Condition' in the control actions. This filters reactions to only flag emojis that represent countries. In the condition builder, set the left side to 'Reaction' from the dynamic content menu. Set the operator to 'contains' and the right side to 'flag-' since Slack flag emojis follow the pattern ':flag-us:', ':flag-es:', etc.

1. 1Click '+ New step' under the Slack trigger
2. 2Search for 'Condition' and select it
3. 3Click in the left condition box
4. 4Select 'Reaction' from dynamic content
5. 5Set operator to 'contains'
6. 6Type 'flag-' in the right value box

Condition Yes branch > Data Operation > Compose

Map flag emoji to language code

In the 'Yes' branch of the condition, add a 'Compose' action to convert flag emojis to language codes OpenAI understands. Use a switch statement or nested conditions to map common flags: flag-us and flag-gb to 'English', flag-es to 'Spanish', flag-fr to 'French', flag-de to 'German', flag-it to 'Italian', flag-jp to 'Japanese', flag-kr to 'Korean', flag-cn to 'Chinese'. This ensures OpenAI gets proper language names instead of emoji codes.

1. 1Click 'Add an action' in the Yes branch
2. 2Search for 'Compose' and select it
3. 3Click in the Inputs box
4. 4Build a switch expression using if() functions
5. 5Map each flag- pattern to its language name

Add this expression in the Compose action to handle more flag emoji variations and provide fallback language detection when users react with non-standard country emojis.

if(
  contains(triggerOutputs()?['body/reaction'], 'flag-us'),
  'English',

if(
  contains(triggerOutputs()?['body/reaction'], 'flag-us'),
  'English',
  if(
    contains(triggerOutputs()?['body/reaction'], 'flag-gb'),
    'English',
    if(
      contains(triggerOutputs()?['body/reaction'], 'flag-es'),
      'Spanish',
      if(
        contains(triggerOutputs()?['body/reaction'], 'flag-fr'),
        'French',
        if(
          contains(triggerOutputs()?['body/reaction'], 'flag-de'),
          'German',
          if(
            contains(triggerOutputs()?['body/reaction'], 'flag-it'),
            'Italian',
            if(
              contains(triggerOutputs()?['body/reaction'], 'flag-jp'),
              'Japanese',
              'English'
            )
          )
        )
      )
    )
  )
)

Condition Yes branch > Slack > Get message

Get the original message content

Add another action and search for 'Get message' from the Slack connector. This retrieves the full text of the message that received the flag reaction. Set the Team to your workspace and Channel to the 'Channel' value from the trigger dynamic content. For Message Timestamp, use the 'Message timestamp' field from the trigger - this uniquely identifies which message to translate.

1. 1Click 'Add an action' after the Compose step
2. 2Search for 'Get message' in Slack actions
3. 3Set Team to your workspace
4. 4Set Channel to dynamic 'Channel' from trigger
5. 5Set Message Timestamp to 'Message timestamp' from trigger

Condition Yes branch > OpenAI > Complete chat completion

Connect to OpenAI

Add a new action and search for 'OpenAI' or 'HTTP' if OpenAI connector isn't available. If using the OpenAI connector, select 'Complete chat completion'. If using HTTP, you'll make a POST request to api.openai.com/v1/chat/completions. For direct OpenAI connector, you'll need your API key from platform.openai.com - go to API Keys section and create a new secret key specifically for this automation.

1. 1Click 'Add an action' after Get message
2. 2Search for 'OpenAI' in the connector list
3. 3Select 'Complete chat completion'
4. 4Click 'Sign in' or 'Add connection'
5. 5Paste your OpenAI API key from platform.openai.com

OpenAI action > Configure parameters

Configure translation prompt

Set the Model to 'gpt-3.5-turbo' for cost efficiency or 'gpt-4' for better accuracy. In the Messages array, create a system message with the role 'system' and content like 'You are a translator. Translate the following message to [Language] and return only the translation, no additional text.' For the user message, set role to 'user' and content to the 'Text' field from the Get message action. Replace [Language] with the output from your Compose action.

1. 1Set Model to 'gpt-3.5-turbo'
2. 2Click 'Add new item' in Messages array
3. 3Set Role to 'system' for first message
4. 4Add translation instruction in Content field
5. 5Add second message with Role 'user'
6. 6Set Content to 'Text' from Get message dynamic content

Condition Yes branch > Slack > Post message to a channel

Post translation back to Slack

Add a final Slack action called 'Post message to a channel'. Set the Team to your workspace and Channel to the same 'Channel' from the trigger. For the message text, combine the flag emoji, original user mention, and translation. Use dynamic content to build something like '🇪🇸 Translation for @[User]: [OpenAI response]'. Set 'Post as' to 'Flow bot' so translations appear from your automation.

1. 1Add final action and search for 'Post message'
2. 2Select 'Post message to a channel' from Slack
3. 3Set Team to your workspace
4. 4Set Channel to 'Channel' from trigger
5. 5Build message text with flag, user mention, and translation
6. 6Set 'Post as' to 'Flow bot'

Flow designer > Save

Save and test the flow

Click 'Save' in the top right corner and wait for validation to complete. Go to a Slack channel and post a test message, then react to it with a flag emoji like 🇪🇸. Watch the flow run history in Power Automate - successful runs show green checkmarks for each step. The translation should appear in Slack within 10-15 seconds if everything is configured correctly.

1. 1Click 'Save' button in top navigation
2. 2Wait for 'Your flow is ready to go' confirmation
3. 3Go to your Slack workspace
4. 4Post a test message in any channel
5. 5React with a flag emoji
6. 6Check Power Automate run history for results