Struct HybridSearchItem 

pub struct HybridSearchItem {

    pub memory_id: i64,
    pub name: String,
    pub namespace: String,
    pub memory_type: String,
    pub description: String,
    pub body: String,
    pub snippet: String,
    pub combined_score: f64,
    pub score: f64,
    pub source: String,
    pub vec_rank: Option<usize>,
    pub fts_rank: Option<usize>,
    pub rrf_score: Option<f64>,
    pub normalized_score: f64,
    pub vec_distance: Option<f64>,
    pub fts_bm25: Option<f64>,
}

Hybrid search item.

Fields

memory_id: i64

Memory identifier.

name: String

Name of this item.

namespace: String

Namespace scope.

memory_type: String

Memory type classification.

description: String

Human-readable description.

body: String

Full text body.

snippet: String

Snippet.

combined_score: f64

Combined score.

score: f64

Alias of combined_score for the documented contract in SKILL.md.

source: String

Source of the match: always “hybrid” (RRF of vec + fts). Added in v2.0.1.

vec_rank: Option<usize>

VEC rank.

fts_rank: Option<usize>

FTS rank.

rrf_score: Option<f64>

Combined RRF score — explicit alias of combined_score for integration contracts.

normalized_score: f64

RRF score normalized to [0.0, 1.0] for cross-method comparability.

vec_distance: Option<f64>

Raw KNN distance from the vector index (lower = more similar).

Present when the result came from the vector search path; None when the result appeared only in the FTS5 results and was not ranked by the KNN index.

fts_bm25: Option<f64>

Raw BM25 score from the FTS5 index. Currently always None; reserved for a future release when the FTS5 BM25 score is exposed by the storage layer.

Trait Implementations

impl Serialize for HybridSearchItem

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.