Sort incoming email into sales, support, billing, spam or newsletter, and flag urgent messages.

• categorychoice

  Which inbox should this email go to?

  salessupportbillingspamnewsletterother
• urgentyes / no

  Does this email need a reply today?

template: email-triage · 1 credit / run

A single yes/no gate for any message: is it spam?

• is_spamyes / no

  Is this message spam?

template: spam-detection · 1 credit / run

Decide whether an inbound lead is qualified, score its fit, and tag industry and buying intent.

• qualifiedyes / no

  Is this lead a real business prospect worth a sales follow-up?

• fit_scorescore

  How well does this lead fit a B2B software product?

  5 levels: No fit: personal use or unrelated need → Excellent fit: urgent need, established company

• industrychoice

  Which industry is the lead's company in?

  softwareecommercefinancehealthcareagencyother
• intentchoice

  What is the lead trying to do right now?

  buy_nowevaluatingresearchingnone

template: lead-qualification · 1 credit / run

Categorize support tickets, score urgency and sentiment, and flag tickets that need escalation.

• categorychoice

  What is this support ticket about?

  bugbillingaccounthow_tofeature_requestother
• urgencyscore

  How urgent is this ticket for the customer?

  4 levels: Low: a question or cosmetic issue → Critical: outage, data loss or security issue

• sentimentscore

  How does the customer feel?

  4 levels: Angry → Positive

• escalateyes / no

  Should this ticket be escalated to a senior engineer or manager?

template: support-triage · 1 credit / run

Route website contact form submissions to sales, support, partnerships or hiring, and drop spam.

• routechoice

  Which team should receive this contact form submission?

  salessupportpartnershipjob_applicationspam

template: contact-form-routing · 1 credit / run

Check an AI agent's proposed action before it runs: allow it, block it, or send it to a human.

• verdictchoice

  Should the agent be allowed to perform the proposed action for this task?

  allowreviewreject
• riskscore

  How risky is the proposed action?

  5 levels: None → Severe

template: agent-action-guardrail · 1 credit / run

Identify what kind of document a text is before routing it to the right workflow.

• document_typechoice

  What type of document is this?

  invoicereceiptcontractresumepurchase_orderother

template: document-classification · 1 credit / run