AxPlaybookOptions

type AxPlaybookOptions = object;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L21

Options for playbook.

A playbook grows an evolving body of task knowledge (“context engineering”) and renders it into a program’s context. The underlying evolution engine is an implementation detail (currently ACE — the Agentic Context Engineering loop) and is intentionally absent from this surface, mirroring how optimize hides its optimizer.

Properties

allowDynamicSections?

optional allowDynamicSections: boolean;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L41

Allow the playbook to grow new sections on its own.

auto?

optional auto: "light" | "medium" | "heavy";

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L45

Intensity preset applied at construction.

initialPlaybook?

optional initialPlaybook: AxACEPlaybook;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L43

Seed the playbook with existing content.

maxEpochs?

optional maxEpochs: number;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L35

Max passes over the dataset during AxPlaybook.evolve.

maxReflectorRounds?

optional maxReflectorRounds: number;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L37

Max reflection refinement rounds per example.

maxSectionSize?

optional maxSectionSize: number;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L39

Max bullets per section before pruning kicks in.

seed?

optional seed: number;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L33

studentAI

studentAI: AxAIService;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L23

Model that runs the program while the playbook is grown.

teacherAI?

optional teacherAI: AxAIService;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L25

Model used to reflect on rollouts and curate the playbook. Defaults to studentAI.

teacherOptions?

optional teacherOptions: AxAIServiceOptions;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L31

AI service options for the reflection and curation calls, including when they fall back to studentAI. Set useExpensiveModel: 'yes' to use a teacher model marked isExpensive.

verbose?

optional verbose: boolean;

Defined in: https://github.com/ax-llm/ax/blob/3d5aa8dac28a2637e4150f022dcd02e6fc3309f0/src/ax/dsp/playbook.ts#L32