AxAgentPlaybookOptions

TypeScript

type AxAgentPlaybookOptions = object & Pick<AxPlaybookOptions, 
  | "verbose"
  | "seed"
  | "maxEpochs"
  | "maxReflectorRounds"
  | "maxSectionSize"
  | "allowDynamicSections"
  | "initialPlaybook"
| "auto">;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/agent/agentInternal/agentOptimizeTypes.ts#L187

Options for AxAgent.playbook(). Builds an AxPlaybook bound to an agent stage (the actor by default). The evolution engine (ACE) is hidden, exactly as in the standalone playbook() factory.

Type Declaration

Name	Type	Description
`apply?`	`boolean`	Render the evolving playbook into the live stage. Defaults to `true`.
`studentAI?`	`Readonly`<`AxAIService`>	-
`target?`	`"actor"` \| `"responder"`	Which agent stage to evolve a playbook for. Defaults to `'actor'`.
`teacherAI?`	`Readonly`<`AxAIService`>	-
`teacherOptions?`	`AxAIServiceOptions`	AI service options for the playbook’s reflection and curation calls, which use `teacherAI`, then the agent’s `judgeAI`, then the student model. Set `useExpensiveModel: 'yes'` for a teacher marked `isExpensive`.