How to Create Thread Summaries with OpenAI and Slack via N8n

Workflow Canvas > + > Slack Trigger

Add Slack Trigger Node

Set up N8n to listen for new emoji reactions in your Slack workspace. This will fire every time someone adds any emoji to a message.

1. 1Click the + button to add a new node
2. 2Search for 'Slack' and select 'Slack Trigger'
3. 3Choose 'On Reaction Added' from the event dropdown
4. 4Click 'Create New Credential' to connect your Slack workspace

Slack Trigger > Credentials > OAuth2 API

Connect Slack Workspace

Authenticate N8n with your Slack workspace using OAuth. This requires admin permissions to install the N8n bot.

1. 1Click 'OAuth2 API' as the authentication method
2. 2Click 'Connect my account' to open Slack OAuth
3. 3Select your workspace and click 'Allow'
4. 4Return to N8n and click 'Save' on the credential

Workflow Canvas > + > IF

Filter for Notebook Emoji

Add a filter to only process 📝 reactions, ignoring all other emoji. This prevents the workflow from firing on every thumbs up or checkmark.

1. 1Add an IF node after the Slack Trigger
2. 2Set condition to 'String' and 'Equal'
3. 3In Value 1, select 'reaction' from the Slack trigger data
4. 4In Value 2, type 'memo' (the emoji shortcode for 📝)

Workflow Canvas > + > Slack > Get Channel Messages

Get Thread Messages

Fetch all messages in the thread using Slack's conversations.replies API. This pulls the original message plus all replies in chronological order.

1. 1Connect a Slack node to the IF node's True branch
2. 2Set operation to 'Get Channel Messages'
3. 3Map 'Channel' to the channel ID from the trigger
4. 4Map 'Timestamp' to the message timestamp from trigger data
5. 5Enable 'Return All' to get the complete thread

Workflow Canvas > + > Function

Format Thread Text

Combine all thread messages into a single text block for OpenAI. Add timestamps and user names to preserve context in the summary.

1. 1Add a Function node after the Slack messages node
2. 2Select 'Run Once for All Items' mode
3. 3Paste the thread formatting code into the code editor
4. 4Test the function to verify proper text concatenation

Workflow Canvas > + > OpenAI > Credentials

Connect OpenAI API

Set up OpenAI credentials to access GPT models for summarization. You'll need an active OpenAI API account with available credits.

1. 1Add an OpenAI node after the Function node
2. 2Click 'Create New Credential' for OpenAI
3. 3Paste your OpenAI API key from platform.openai.com
4. 4Click 'Save' and test the connection

OpenAI > Chat > Message a model

Configure GPT Summarization

Set up the OpenAI node to summarize thread content using GPT-3.5 or GPT-4. Include specific instructions for thread context and summary length.

1. 1Set Resource to 'Chat' and Operation to 'Message a model'
2. 2Choose 'gpt-3.5-turbo' or 'gpt-4' as the model
3. 3Set the system message: 'Summarize this Slack thread in 2-3 sentences'
4. 4Map the user message to the formatted thread text from Function node
5. 5Set max tokens to 150 for concise summaries

Workflow Canvas > + > Slack > Post Message

Post Summary to Thread

Send the AI summary back to the original Slack thread as a reply. This keeps the summary contextually linked to the conversation.

1. 1Add another Slack node after OpenAI
2. 2Set operation to 'Post Message'
3. 3Map Channel to the original channel from trigger data
4. 4Set Thread Timestamp to the original message timestamp
5. 5Map message text to OpenAI's response content
6. 6Add a prefix like '📝 Thread Summary: '

OpenAI Node > Settings > Continue on Fail

Add Error Handling

Configure the workflow to handle API failures gracefully. This prevents broken executions when OpenAI is down or rate-limited.

1. 1Click the OpenAI node settings (gear icon)
2. 2Set 'On Error' to 'Continue (using Error Output)'
3. 3Connect the error output to a Slack node
4. 4Configure error message: 'Sorry, couldn't summarize this thread right now. Try again in a few minutes.'

Workflow Header > Active Toggle

Test and Activate

Test the complete workflow with a real Slack thread and deploy it for your team. Start with a test channel before rolling out workspace-wide.

1. 1Click 'Save' to save your workflow
2. 2Toggle the workflow Active switch to On
3. 3Go to a Slack thread and react with 📝
4. 4Check N8n executions tab for successful completion
5. 5Verify the summary appears in the Slack thread

Drop this into an n8n Code node.

// Function node code to clean and format thread messages
const messages = $input.all();
const formatted = messages

// Function node code to clean and format thread messages
const messages = $input.all();
const formatted = messages
  .filter(msg => msg.json.user && !msg.json.bot_id)
  .map(msg => `${msg.json.user}: ${msg.json.text}`)
  .join('\n');

return [{ json: { thread_text: formatted } }];