AxAgentPlaybook
Type Parameters
| Type Parameter | Default type |
|---|---|
IN extends AxGenIn | AxGenIn |
OUT extends AxGenOut | AxGenOut |
Constructors
Constructor
new AxAgentPlaybook<IN, OUT>(agent: unknown, handle: AxPlaybook<IN, OUT>): AxAgentPlaybook<IN, OUT>;Parameters
| Parameter | Type | Description |
|---|---|---|
agent | unknown | The owning agent coordinator (used by evolve). |
handle | AxPlaybook<IN, OUT> | The inner stage-bound playbook the handle methods delegate to. |
Returns
AxAgentPlaybook<IN, OUT>
Accessors
inner
Get Signature
get inner(): AxPlaybook<IN, OUT>;The inner program-level playbook handle (for advanced use).
Returns
AxPlaybook<IN, OUT>
Methods
configureAuto()
configureAuto(level: "medium" | "light" | "heavy"): void;Set the evolution intensity preset.
Parameters
| Parameter | Type |
|---|---|
level | "medium" | "light" | "heavy" |
Returns
void
evolve()
evolve(dataset: Readonly<AxAgentEvalDataset<IN>>, options?: Readonly<AxAgentPlaybookEvolveOptions>): Promise<AxAgentPlaybookEvolveResult<OUT>>;Grow the playbook from a task set. verify (default) keeps only bullets
that provably help — re-scoring train + held-out after each candidate and
rolling back regressions. verify: false applies mined lessons without
the gate (trust-batch). Produces only playbook bullets. Must not run
concurrently with forward() on the same agent instance.
Parameters
| Parameter | Type |
|---|---|
dataset | Readonly<AxAgentEvalDataset<IN>> |
options? | Readonly<AxAgentPlaybookEvolveOptions> |
Returns
Promise<AxAgentPlaybookEvolveResult<OUT>>
getState()
getState(): AxPlaybookSnapshot;A serializable snapshot of the current playbook and its history.
Returns
load()
load(snapshot: Readonly<AxPlaybookSnapshot>): this;Restore a snapshot and render it into the live stage.
Parameters
| Parameter | Type |
|---|---|
snapshot | Readonly<AxPlaybookSnapshot> |
Returns
this
render()
render(): string;The current playbook rendered as a markdown block.
Returns
string
reset()
reset(): void;Clear the playbook back to its initial state.
Returns
void
toJSON()
toJSON(): AxPlaybookSnapshot;Alias of getState so JSON.stringify(handle) yields a snapshot.
Returns
update()
update(args: Readonly<{
example: unknown;
feedback?: string;
prediction: unknown;
}>): Promise<void>;Refine the playbook from a single live interaction (trust).
Parameters
| Parameter | Type |
|---|---|
args | Readonly<{ example: unknown; feedback?: string; prediction: unknown; }> |
Returns
Promise<void>