Crate vipune 

vipune - A minimal memory layer for AI agents.

This crate provides a local, semantic memory store with conflict detection. All operations are synchronous (no async/await required).

Example

use vipune::{Config, MemoryStore, detect_project};

// Initialize memory store
let config = Config::default();
let mut store = MemoryStore::new(
    config.database_path.as_path(),
    &config.embedding_model,
    config.clone()
).expect("Failed to initialize store");

// Detect project ID
let project_id = detect_project(None);

// Add a memory with conflict detection
let result = store.add_with_conflict(&project_id, "Alice works at Microsoft", None, false);
match result {
    Ok(vipune::AddResult::Added { id }) => println!("Added memory: {}", id),
    Ok(vipune::AddResult::Conflicts { .. }) => println!("Conflict detected"),
    Err(e) => eprintln!("Error: {}", e),
    Err(e) => eprintln!("Error: {}", e),
}

// Search memories
let results = store.search(&project_id, "where does alice work", 10, 0.0);
for memory in results.unwrap() {
    println!("{:.2}: {}", memory.similarity.unwrap_or(0.0), memory.content);
}

Mutability Requirements

Methods that generate embeddings (add, search, update) require &mut self because the embedding engine internally mutates state for ONNX tensor allocations.

Re-exports

pub use config::Config;

pub use embedding::EMBEDDING_DIMS;

pub use embedding::EmbeddingEngine;

pub use errors::Error;

pub use memory::MemoryStore;

pub use memory_types::AddResult;

pub use memory_types::ConflictMemory;

pub use project::detect_project;

Modules

config

Configuration system for vipune.

embedding

Synchronous ONNX embedding engine for text-to-vector conversion.

errors

Error types for vipune.

memory

Core memory store orchestrating embedding and SQLite operations.

memory_types

Memory store data types.

project

Project auto-detection from git repository

Structs

Memory

A single memory record with metadata and optional similarity score.

Constants

MAX_INPUT_LENGTH

Maximum allowed input length (100,000 characters).

MAX_SEARCH_LIMIT

Maximum allowed limit for search operations.