Front and Slack: AI Conversation-Tag Suggester Template

When a new Front conversation arrives, an Agent-mode Connector node picks the best tags from your taxonomy as structured output, applies them in Front, and notes the choice in Slack.

What It Builds

This Spojit template starts from a Webhook trigger that Front calls when a new conversation lands in an inbox. An Agent-mode Connector node on the front connector reads the conversation, compares it against your existing tag list, and returns its choices as structured JSON. The workflow then applies those tags back onto the conversation with the front connector and posts a short note to a slack channel so your team can see what was tagged and why.

The result is consistent, hands-off triage: every incoming conversation gets labeled against the same taxonomy the moment it arrives, and your team gets a running feed of tagging decisions in Slack.

The Prompt

Paste this into Miraxa, the intelligent layer across your automation, and it builds the workflow, connecting the tools for you:

Build a workflow triggered by a Front webhook when a new conversation arrives. Use the Front connector to read the conversation and to list the existing tags in the inbox. Then use a Front connector node in Agent mode to suggest the one or two best-matching tags from that list, returning them as structured output. Apply the chosen tags to the conversation in Front, then post a message to the #support-triage Slack channel with the conversation subject, a short link reference, and the tags that were applied.

Connectors Used

front - Webhook trigger fires on a new conversation; reads the conversation, lists your existing tags, and applies the chosen tags with add-tag.
slack - posts the tagging summary to a channel with send-message.

Customize It

Change #support-triage to the channel your team watches, and adjust "one or two best-matching tags" to allow more or fewer labels. To bias the suggestions, add a sentence describing your taxonomy (for example, "prefer the billing, bug, and feature-request tags") so the agent leans toward the categories that matter most. You can also widen the trigger to a specific inbox by configuring the Front webhook to send only that inbox's events.

Tips

Let the agent list your existing tags first so it only ever suggests labels that already exist in Front, rather than inventing new ones.
Use a Response Schema on the Agent-mode Connector node to force the output into a clean JSON array of tag names, which makes the apply step reliable.
Keep the actual tag-apply and Slack steps in Direct mode: they are single, predictable calls and run without AI credits.

Common Pitfalls

If the Front Webhook trigger never fires, confirm the webhook is registered in Front for the right inbox and verified by the matching signing connection in Spojit.
The agent can only apply tags that exist; if a suggested label is missing, the apply step will fail, so keep your taxonomy in sync with what the prompt expects.
High-volume inboxes can generate a burst of runs, so watch your Slack channel for noise and narrow the trigger to one inbox if needed.