Skip to main content

​
Overview

The Memory layer (Layer 2) provides persistent storage and retrieval of information with metadata support. It implements multi-dimensional search across keyword, temporal, and semantic dimensions.

​
Memory Trait

The Memory trait defines the core interface for storing and retrieving memory entries. 
pub trait Memory: Send + Sync {
    fn store(&self, content: &str, meta: MemoryMeta) -> Result<String>;
    fn get(&self, id: &str) -> Result<Option<MemoryEntry>>;
    fn search(&self, query: &MemoryQuery) -> Result<Vec<MemoryEntry>>;
    fn delete(&self, id: &str) -> Result<bool>;
    fn count(&self) -> Result<usize>;
    fn as_vector(&self) -> Option<&dyn VectorMemory>;
}

​
Methods

​
store()

fn store(&self, content: &str, meta: MemoryMeta) -> Result<String>
Stores content with metadata and returns a unique entry ID. Parameters:
  • content: The text content to store
  • meta: Metadata including tags, priority, and source
Returns: Entry ID (UUID string) Source: traits.rs:112

​
get()

fn get(&self, id: &str) -> Result<Option<MemoryEntry>>
Retrieves a memory entry by its unique identifier. Parameters:
  • id: The entry identifier
Returns: Some(MemoryEntry) if found, None otherwise Source: traits.rs:115 
fn search(&self, query: &MemoryQuery) -> Result<Vec<MemoryEntry>>
Performs multi-dimensional search across text, tags, time range, and priority. Parameters:
  • query: Search query with filters
Returns: Vector of matching entries Source: traits.rs:118

​
delete()

fn delete(&self, id: &str) -> Result<bool>
Deletes an entry by ID. Parameters:
  • id: The entry identifier
Returns: true if entry was deleted, false if not found Source: traits.rs:121

​
count()

fn count(&self) -> Result<usize>
Returns the total number of stored entries. Source: traits.rs:124

​
as_vector()

fn as_vector(&self) -> Option<&dyn VectorMemory>
Upcasts to VectorMemory if the implementation supports vector search. Returns None by default. Source: traits.rs:128

​
VectorMemory Trait

Extends Memory with vector embedding and semantic search capabilities. 
pub trait VectorMemory: Memory {
    fn store_with_embedding(&self, content: &str, meta: MemoryMeta, embedding: &Embedding) -> Result<String>;
    fn vector_search(&self, query: &VectorQuery) -> Result<Vec<VectorSearchResult>>;
    fn hybrid_search(&self, text: &str, query_embedding: &Embedding, limit: usize) -> Result<Vec<VectorSearchResult>>;
    fn has_vector_support(&self) -> bool;
    fn vector_stats(&self) -> Result<VectorStats>;
}
Source: vector.rs:115 See Vector Search API for detailed documentation.

​
Data Types

​
Priority

pub enum Priority {
    Low = 0,
    Medium = 1,    // default
    High = 2,
    Critical = 3,
}
Priority levels for memory entries. Source: traits.rs:11

​
MemoryMeta

pub struct MemoryMeta {
    pub tags: Vec<String>,
    pub priority: Priority,
    pub source: String,
}
Metadata associated with each memory entry. Source: traits.rs:24

​
MemoryEntry

pub struct MemoryEntry {
    pub id: String,
    pub content: String,
    pub meta: MemoryMeta,
    pub created_at: DateTime<Utc>,
    pub updated_at: DateTime<Utc>,
}
A complete memory entry with content, metadata, and timestamps. Source: traits.rs:45

​
MemoryQuery

pub struct MemoryQuery {
    pub text: String,
    pub tags: Vec<String>,
    pub after: Option<DateTime<Utc>>,
    pub before: Option<DateTime<Utc>>,
    pub min_priority: Option<Priority>,
    pub limit: usize,
}
Multi-dimensional search query with builder methods: 
let query = MemoryQuery::new("blood pressure")
    .with_tags(vec!["sensor".into()])
    .with_min_priority(Priority::High)
    .with_time_range(Some(start_time), None)
    .with_limit(20);
Source: traits.rs:60

​
NoopMemory Implementation

In-memory stub implementation for testing. 
pub struct NoopMemory {
    entries: std::sync::Mutex<Vec<MemoryEntry>>,
}

​
Example Usage

use oneclaw_core::memory::{NoopMemory, Memory, MemoryMeta, MemoryQuery};

let mem = NoopMemory::new();

// Store an entry
let id = mem.store("blood pressure 140/90", MemoryMeta::default())?;

// Retrieve by ID
let entry = mem.get(&id)?;

// Search
let results = mem.search(&MemoryQuery::new("blood pressure"))?;
Features:
  • Thread-safe with Mutex
  • No persistence (data lost on drop)
  • Simple substring matching for text search
  • AND logic for tag filtering
Source: traits.rs:134

Build docs developers (and LLMs) love

Get started for freeTalk to us