Agents and extractors

Add an agent when a stage needs model reasoning. Add an extractor when the next stage needs validated fields from existing content.

Prompt an agent

const supportSummary = new PipelineBuilder(TicketInput)
  .step((ticket) => [
    `Customer: ${ticket.customer}`,
    `Subject: ${ticket.subject}`,
    `Body: ${ticket.body}`,
  ].join('\n'))
  .step((text) =>
    `Write a concise internal support summary:\n\n${text}`,
  )
  .prompt(summaryAgent)
  .build()

.prompt(agent) converts the current value with String(value). Format objects explicitly first; otherwise the model may receive [object Object].

Extract typed fields

const triagePipeline = new PipelineBuilder(TicketInput)
  .step((ticket) => `${ticket.subject}\n\n${ticket.body}`)
  .prompt(summaryAgent)
  .extract(ticketExtractor)
  .step((ticket) => ({
    ...ticket,
    route: ticket.priority === 'high' ? 'incident' : 'support',
  }))
  .build()

After .extract(...), the next stage receives the extractor's schema type. Extractor retries and validation behavior remain configured on the extractor.

Keep boundaries clear

The agent owns reasoning. The extractor owns converting text into validated fields. Deterministic steps own authorization, product state, side effects, and final response mapping.