Class OutcomeFeedbackCollector

Outcome feedback collector implementation.

Implements

IOutcomeFeedback

Index

Constructors

Methods

registerLinUCBBandit registerPreferenceRouter onOutcomeProcessed recordRoutingDecision recordOutcome processOutcome computeReward getStats getPendingDecisions clearExpiredDecisions reset

Constructors

constructor

new OutcomeFeedbackCollector(
config?: Partial<FeedbackCollectorConfig>,
): OutcomeFeedbackCollector
Parameters
- Optionalconfig: Partial<FeedbackCollectorConfig>
Returns OutcomeFeedbackCollector
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:55

Methods

registerLinUCBBandit

registerLinUCBBandit(bandit: LinUCBBandit): void
Register LinUCB bandit for direct feedback.
Parameters
- bandit: LinUCBBandit
Returns void
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:70

registerPreferenceRouter

registerPreferenceRouter(router: PreferenceRouter): void
Register PreferenceRouter for direct feedback.
Parameters
- router: PreferenceRouter
Returns void
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:78

onOutcomeProcessed

onOutcomeProcessed(callback: OutcomeProcessedCallback): () => void
Subscribe to outcome processed events.
Parameters
- callback: OutcomeProcessedCallback
Returns () => void
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:86

recordRoutingDecision

recordRoutingDecision(decision: FeedbackRoutingDecision): void
Record a routing decision for tracking.
Parameters
- decision: FeedbackRoutingDecision
Returns void
Implementation of IOutcomeFeedback.recordRoutingDecision
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:93

recordOutcome

recordOutcome(outcome: TaskOutcome): void
Record an outcome for a routing decision.
Parameters
- outcome: TaskOutcome
Returns void
Implementation of IOutcomeFeedback.recordOutcome
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:119

processOutcome

processOutcome(
traceId: string,
partialOutcome: Omit<TaskOutcome, "routingDecisionId">,
): void
Process outcome for a trace ID (finds matching decision and computes reward).
Parameters
- traceId: string
- partialOutcome: Omit<TaskOutcome, "routingDecisionId">
Returns void
Implementation of IOutcomeFeedback.processOutcome
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:140

computeReward

computeReward(outcome: TaskOutcome): ComputedReward
Compute reward from an outcome.
Parameters
- outcome: TaskOutcome
Returns ComputedReward
Implementation of IOutcomeFeedback.computeReward
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:156

getStats

getStats(): FeedbackLoopStats
Get feedback loop statistics.

Returns FeedbackLoopStats
Implementation of IOutcomeFeedback.getStats
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:196

getPendingDecisions

getPendingDecisions(): readonly FeedbackRoutingDecision[]
Get pending decisions (waiting for outcomes).

Returns readonly FeedbackRoutingDecision[]
Implementation of IOutcomeFeedback.getPendingDecisions
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:214

clearExpiredDecisions

clearExpiredDecisions(): number
Clear expired pending decisions.

Returns number
Implementation of IOutcomeFeedback.clearExpiredDecisions
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:218

reset

reset(): void
Reset all state.

Returns void
Implementation of IOutcomeFeedback.reset
- Defined in packages/nexus-agents/src/learning/outcome-feedback.ts:250