Knowledge and Slack: AI Policy-Answer Bot Template

When a teammate asks a policy question from Slack, Spojit answers it from your own policy documents with a cited response and posts the reply straight back into the thread.

What It Builds

This Spojit template starts from a Webhook trigger that receives a Slack-sourced question, verified with a Slack signing connection. A Knowledge node in Query mode searches your policy collection and produces a structured, cited answer drawn only from the documents you embedded. A Connector node on the slack connector then uses send-message to post that answer back into the same Slack thread, so the person who asked sees the response where they asked it.

Because the Knowledge node runs in Query mode, the answer is grounded in your uploaded policy files rather than general knowledge, and the Response Schema keeps the output structured so you can post a clean answer plus its source citations.

The Prompt

Paste this into Miraxa and it builds the workflow, connecting the tools for you:

Build a workflow with a Webhook trigger (verified by a Slack signing connection) that receives a question text and a Slack channel and thread timestamp. Add a Knowledge node in Query mode that queries my "company-policies" collection with the question and returns a structured answer with cited sources. Then add a Slack Connector node in Direct mode that uses send-message to post the answer and its citations back to that channel and thread.

Connectors Used

  • Webhook trigger - receives the incoming Slack question and is verified with a Slack signing connection.
  • Knowledge node (Query mode) - searches your policy collection and returns a cited, structured answer.
  • slack - posts the answer back into the thread with send-message.

Customize It

Change the collection name (company-policies) to whichever Knowledge collection holds your documents, adjust the Result Count in the Knowledge node to pull more or fewer source chunks per answer, and tune the Response Schema if you want extra fields like a confidence note or a "no answer found" flag. You can also swap the channel and thread mapping in the Slack send-message step if you route questions through a single help channel instead of replying in-thread.

Tips

  • Upload and embed your policy files into the collection first, in the Knowledge section of the sidebar, and confirm each document shows status READY before you run the workflow.
  • Keep the Slack step in Direct mode with send-message so posting the reply is deterministic and costs no AI credits; reserve AI spend for the Knowledge query.
  • Use the same embedding model for every document in the collection so queries match all of your content consistently.

Related

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.