Interface CognitiveRetrievalOptions

interface CognitiveRetrievalOptions {
    topK?: number;
    types?: MemoryType[];
    scopes?: {
        scope: MemoryScope;
        scopeId: string;
    }[];
    tags?: string[];
    entities?: string[];
    minConfidence?: number;
    timeRange?: {
        after?: number;
        before?: number;
    };
    neutralMood?: boolean;
    hyde?: boolean;
}

Properties

topK? types? scopes? tags? entities? minConfidence? timeRange? neutralMood? hyde?

Properties

Optional topK

topK?: number

Optional types

types?: MemoryType[]

Optional scopes

scopes?: {
    scope: MemoryScope;
    scopeId: string;
}[]

Type declaration

Optional tags

tags?: string[]

Optional entities

entities?: string[]

Optional minConfidence

minConfidence?: number

Optional timeRange

timeRange?: {
    after?: number;
    before?: number;
}

Type declaration

  • Optional after?: number
  • Optional before?: number

Optional neutralMood

neutralMood?: boolean

If true, skip emotional congruence bias (useful for factual lookups).

Optional hyde

hyde?: boolean

Enable HyDE (Hypothetical Document Embedding) for memory retrieval.

When true and a HyDE retriever is configured on the memory manager, the system generates a hypothetical memory trace matching the query before embedding. This produces embeddings that are closer to actual stored memories, improving recall — especially for vague or abstract recall prompts (e.g. "that thing we discussed about deployment").

Adds one LLM call per retrieval. Use for important lookups where recall quality matters more than latency.

Default

false

Settings

Member Visibility

Theme

On This Page

topKtypesscopestagsentitiesminConfidencetimeRangeneutralMoodhyde