Interface: LessonScoringBuilder

Defined in: builders/lesson-scoring-builder.ts:9

Fluent configurator for lesson scoring and recall parameters ($s in .scoring($s => ...)).

Configures the sliding-window injection count $K$, partition match boost $\beta$, epoch decay factor $\gamma$, and recall filtering bounds (minScore and maxItems).

Methods

build()

build(): LessonScoringConfig;

Defined in: builders/lesson-scoring-builder.ts:61

Compiles and returns the frozen LessonScoringConfig.

Freezes all scoring parameters and nested recall bounds into a passive AST record.

Returns

LessonScoringConfig

Deeply frozen scoring configuration.

---

decayOnPartitionEpochChange()

decayOnPartitionEpochChange(decay): LessonScoringBuilder;

Defined in: builders/lesson-scoring-builder.ts:41

Sets the exponential decay factor $\gamma \in (0, 1]$ per partition contract epoch increment.

Attenuates older lessons when the topology contract of their associated partition evolves.

Parameters

Parameter	Type	Description
decay	number	Decay multiplier $\gamma \in (0, 1]$.

Returns

LessonScoringBuilder

New immutable scoring builder.

---

partitionMatchBoost()

partitionMatchBoost(boost): LessonScoringBuilder;

Defined in: builders/lesson-scoring-builder.ts:30

Sets the partition match boost factor $\beta$ applied when a lesson's partitions overlap active partitions.

Increases the composite recall score of partition-specific lessons relative to generic lessons.

Parameters

Parameter	Type	Description
boost	number	Non-negative partition match boost factor $\beta$.

Returns

LessonScoringBuilder

New immutable scoring builder.

---

recall()

recall(opts): LessonScoringBuilder;

Defined in: builders/lesson-scoring-builder.ts:52

Sets the recall filter thresholds (minScore and maxItems) applied at Turn 0.

Only lessons scoring at least opts.minScore up to opts.maxItems are injected into context.

Parameters

Parameter	Type	Description
opts	{ maxItems: number; minScore: number; }	Recall filter bounds.
opts.maxItems	number	-
opts.minScore	number	-

Returns

LessonScoringBuilder

New immutable scoring builder.

---

slidingWindowInjections()

slidingWindowInjections(count): LessonScoringBuilder;

Defined in: builders/lesson-scoring-builder.ts:19

Sets the sliding window size $K$ of deduplicated step executions used for utility scoring.

Determines how many recent injections are evaluated when computing a lesson's empirical pass rate.

Parameters

Parameter	Type	Description
count	number	Positive integer sliding window size $K$.

Returns

LessonScoringBuilder

New immutable scoring builder.