Blocks

Classify

The AI Agent's classify task: define labels, the model picks exactly one — the standard front door to a branch.

Define the labels; the model picks exactly one. Classify is the standard front door to a branch: AI decides the category, the switch routes on it.

Config

The classify task in the Inspector Labels, and optionally what to read and extra guidance.

  • Labels (required) — comma-separated (refund, complaint, question, other). The model must pick one of these — it can't invent a category, which is what makes the output safe to branch on.
  • Input — optional template picking what to read ({{ payload.text }}). Blank reads the whole upstream payload.
  • Extra instructions — optional domain guidance, e.g. "urgent means a same-day deadline".

Output

Structured, always the same two fields:

{ "ok": true,
  "data": { "label": "refund", "reasoning": "Asks for money back on order #2041." },
  "": "…" }

Branch on {{ payload.data.label }}; log or show reasoning when you want the why.

Patterns

  • Ticket triage — classify (urgent, routine, spam) → switch on the label → escalation path per case.
  • Lead routing — classify inbound email by product line → notify the right channel.

Gotchas

  • Always include an escape label (other, unknown) — without one, the model is forced to shoehorn genuinely unclassifiable input into your real categories.
  • Wire every label to a switch case (or its default port) — an unwired label is a run that stops there.