🧠 Mnemos

Memory for Agentic AI. Inspired by minds. Engineered for longevity.

Mnemos is a lightweight, extensible memory system for agentic AI.
It helps agents remember what matters — not just store data.

✨ What is Mnemos?

Mnemos is a memory-as-a-service toolkit that enables developers to add persistent, structured, and coherent memory to AI agents and LLM apps.

Think of it like a hippocampus for your agents:

• Clean API for memory storage and recall
• Optimized for simplicity and developer experience
• Designed for extensibility with pluggable storage backends

🧪 Example Usage

import mnemos

# Store a memory
mnemos.remember("The user prefers minimalist interfaces.", tags=["ui", "preference"])

# Recall related memories
results = mnemos.recall("user interface")
print(results[0].text)  # "The user prefers minimalist interfaces."

🧱 Key Features

• 🧠 Simple, intuitive API with remember() and recall()
• 🔍 Basic text and tag-based search
• 🧪 Fully typed with Python type hints
• 🧰 Extensible storage backends (in-memory included)

📦 Installation

Install with pip:

pip install mnemos

🧪 Running Tests

pytest tests/

🌱 Project Status

Mnemos is in early development. This initial version provides an in-memory implementation with a clean API. Future versions will add persistent storage and more advanced search capabilities.

Contributions, ideas, and feedback are welcome at github.com/iteebz/mnemos

📜 License

MIT