MemNexus Documentation
Persistent memory and behavioral learning for AI assistants.
Welcome to the MemNexus documentation. MemNexus gives your AI assistants persistent memory that works across every platform — Claude, ChatGPT, Cursor, and more.
MemNexus is in invite-only preview. Request access to get your API key — setup takes under 5 minutes.
Get Started
New to MemNexus? Start here.
What is MemNexus?
Understand the problem MemNexus solves and how it works.
Quick Start
Create your first memory in under 5 minutes.
Core Concepts
Memories, conversations, topics, facts, and entities.
How It Works
Architecture and search pipeline at a glance.
Guides
Step-by-step instructions for common tasks.
Using the CLI
Install and use the mx command-line tool.
Using the SDK
Integrate MemNexus into your TypeScript applications.
Integrations
Connect MemNexus to Claude, Cursor, Windsurf, and more.
Workflows
Patterns for getting the most out of persistent memory day-to-day.
Session Start Ritual
Load context in 30 seconds so your AI walks in knowing what matters.
Debugging With Memory
Recognize patterns across incidents and stop investigating the same bug twice.
Team Workflows
Standups, handoffs, and shared context for engineering teams.
Capturing Memories
What to save, how to write it, and when to save it.
AI Agents
Build AI agents with persistent memory.
Overview
Why memory matters for AI agents.
Prompt Library
Pre-built system prompts you can copy and use.
Agent Patterns
Reusable design patterns for memory-aware agents.
Concepts
Deep dives into how MemNexus works under the hood.
Memory Types
Episodic, semantic, and procedural memory — what each stores and when to use it.
Semantic Search
How vector search and graph traversal find relevant memories.
Knowledge Graph
How memories, entities, facts, and topics connect in the graph.
Privacy & Security
Data isolation, encryption, and retention policies.
Reference
Exhaustive technical reference.
CLI Reference
Every mx command documented.
API Reference
OpenAPI 3.0 REST API specification.
Data Model
Schemas, fields, and relationships.