Module objectives 

Retrieval Objective implementations for khive-runtime.

Domain-specific objectives that operate on pre-computed retrieval signals. Pure math: no IO, no async. The runtime layer materialises the signal data and feeds it in via the candidate struct.

See ADR-061 — Retrieval Infrastructure. See ADR-033 — Recall Pipeline (NoteCandidate, DecayAwareSalienceObjective, TemporalRecencyObjective, RerankerObjective).

Structs

AmplifiedDecayAwareSalienceObjective

Scores a NoteCandidate by salience with exponential decay and a non-linear amplification exponent applied after decay.

DecayAwareSalienceObjective

Scores a NoteCandidate by salience with configurable temporal decay.

GraphProximityObjective

Scores a candidate by graph proximity to anchor nodes.

MemoryRecallPipeline

Composable scoring pipeline for memory recall candidates.

NoteCandidate

Pre-computed signals for a single memory note candidate.

RerankerObjective

Scores a NoteCandidate using a named reranker’s pre-computed score.

RetrievalCandidate

Pre-computed retrieval signals for a single candidate entity.

RrfFusionObjective

Scores a candidate by its pre-computed RRF fusion score.

TemporalRecencyObjective

Scores a NoteCandidate by pure temporal recency with a configurable half-life.

TextRelevanceObjective

Scores a candidate by BM25/FTS relevance.

VectorSimilarityObjective

Scores a candidate by cosine similarity to the query vector.