First Steps
A hands-on walkthrough of Mnemosyne's core operations.
1. Initialize Memory
from mnemosyne import Mnemosyne
mem = Mnemosyne()
print(f"Initialized: {mem.db_path}")
2. Store Information
# Simple memory
mem.remember("Alice is a backend engineer who prefers Go over Python.")
# Structured memory with metadata
mem.remember(
content="Project deadline moved to Friday.",
metadata={"project": "alpha", "type": "deadline"},
importance=0.8,
source="slack",
)
3. Recall Information
# Semantic search
results = mem.recall("What programming language does Alice prefer?")
# Filtered search by topic and date
results = mem.recall(
query="deadline",
topic="project-alpha",
from_date="2026-04-01",
)
# Access a specific memory from results
for r in results:
print(f"{r['score']:.2f}: {r['content']}")
4. Update and Delete
# Update a memory
mem.update("mem_abc123", content="Deadline moved to Monday.")
# Delete a memory
mem.forget("mem_abc123")
# No bulk delete — iterate through results
results = mem.recall("temp notes", top_k=10)
for r in results:
mem.forget(r["memory_id"])
5. Inspect Memory State
# Statistics
stats = mem.get_stats()
print(f"Total: {stats['total_memories']} memories across {stats['total_sessions']} sessions")
print(f"BEAM working: {stats['beam']['working_memory']}")
print(f"BEAM episodic: {stats['beam']['episodic_memory']}")
print(f"Triples: {stats['beam']['triples']['total']}")
print(f"Database: {stats['database']}")
You're Ready
You've learned the core CRUD operations. Next, explore the BEAM Architecture to understand how Mnemosyne organizes memory.
Related Pages
Quick Start
Install Mnemosyne and perform your first memory operations in under 5 minutes. Covers pip install, c...
BEAM Architecture
Overview of BEAM (Biological-inspired Episodic-Associative Memory), Mnemosyne's four-tier architectu...
Working Memory
Explore Mnemosyne's Working Memory tier: a short-term, high-speed buffer for the most recent agent o...
Mnemosyne