Struct MemoryQuery 

pub struct MemoryQuery {
    pub limit: usize,
    pub token_budget: Option<usize>,
    pub agent: Option<AgentId>,
    pub semantic: Option<String>,
    pub strategy: RecallStrategy,
}

How to recall: a result limit, an optional agent filter, and an optional semantic query.

Fields

limit: usize

Max items to return.

token_budget: Option<usize>

An optional context-window budget in tokens, alongside the item-count limit. The recall result is rendered to fit it by to_context_block, which fills in rank order until the budget is reached. Advisory: tokens are estimated, not tokenized (no model dependency in the SDK).

agent: Option<AgentId>

semantic: Option<String>

A semantic query. Backends that embed (e.g. VectorMemory) rank results by similarity to this text. The log-backed default ignores it.

strategy: RecallStrategy

How recall finds items. Auto (the default) lets the backend route from the available state: a backend with an embedder and a semantic query ranks by similarity, otherwise it returns the most recent.

Implementations

impl MemoryQuery

pub fn builder() -> MemoryQueryBuilder

Create an instance of MemoryQuery using the builder syntax

Trait Implementations

impl Clone for MemoryQuery

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for MemoryQuery

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.