ActionAgenticSetFieldValues

Per-action config for the "Research & Set Field Values" action. The customer writes a free-text instruction and defines typed outputs; an LLM agent — allowed to evaluate the Contact's Paminga context and/or browse the account's website — fills each output, which is written to its mapped custom field.

type ActionAgenticSetFieldValues {
  browse_website: Boolean!
  created_at: DateTime
  created_by: User
  evaluate_contact: Boolean!
  id: ID!
  instruction: String!
  max_steps: Int
  outputs: JSON
  refresh_after_days: Int
  updated_at: DateTime
  updated_by: User
}

Fields

ActionAgenticSetFieldValues.browse_website ● Boolean! non-null scalar

Whether the agent may browse the account's website (when the domain can be determined).

ActionAgenticSetFieldValues.created_at ● DateTime scalar

ActionAgenticSetFieldValues.created_by ● User object

ActionAgenticSetFieldValues.evaluate_contact ● Boolean! non-null scalar

Whether the agent may read this Contact's Paminga engagement, history & account.

ActionAgenticSetFieldValues.id ● ID! non-null scalar

ActionAgenticSetFieldValues.instruction ● String! non-null scalar

The customer's free-text instruction (the prompt).

ActionAgenticSetFieldValues.max_steps ● Int scalar

Per-run step ceiling (null = the agent's default cap).

ActionAgenticSetFieldValues.outputs ● JSON scalar

The typed outputs to fill: a list of { object_type_id, field_id, guidance }.

ActionAgenticSetFieldValues.refresh_after_days ● Int scalar

Re-run a Contact once its result is older than this many days (null = run once).

ActionAgenticSetFieldValues.updated_at ● DateTime scalar

ActionAgenticSetFieldValues.updated_by ● User object

Member Of

Action object