memory

Manage AI agent memory with notes, MongoDB RAG, and Neo4j graphs.

234|15|Updated Mar 16, 2024
One-click install
npx skills add https://github.com/OriNachum/autonomous-intelligence --skill memory-orinachum
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/OriNachum/autonomous-intelligence/tree/main/qq/.claude/skills-test/memory
Command: npx skills add https://github.com/OriNachum/autonomous-intelligence --skill memory-orinachum

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill provides a robust, multi-layered memory system to ensure that important information is never lost and can be efficiently recalled and utilized.

Core Features & Use Cases

  • Persistent Knowledge: Stores core identity, project details, and system facts that are never forgotten.
  • Contextual Recall: Manages ephemeral working notes, searchable RAG notes via MongoDB, and a structured Neo4j knowledge graph.
  • File Analysis: Extracts knowledge from documents and integrates it across all memory layers.
  • Use Case: An AI assistant can use this Skill to remember user preferences, project status, key entities discussed in conversations, and details from analyzed documents, providing a consistent and informed interaction.

Quick Start

Use the memory skill to analyze the file 'project_plan.md'.

Frequently Asked Questions about memory

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement persistent memory and knowledge graph relationships for an AI agent?

Persistent memory for AI agents is implemented using a multi-layer system combining MongoDB vector search for RAG recall and a Neo4j knowledge graph for entity relationships. This stores core facts and working notes to ensure information is never lost.

What's the best way to extract knowledge from documents and store it in a vector database?

Knowledge extraction from documents is handled through deep file analysis that processes files and automatically integrates the extracted knowledge across all memory layers, including searchable MongoDB RAG vector storage and Neo4j entity graphs.

Do I need Docker to set up MongoDB, Neo4j, and TEI embeddings for RAG applications?

Docker is required to initialize the memory services. The system uses docker-compose to handle service initialization and health checks for MongoDB, Neo4j, and TEI embeddings, ensuring the RAG vector search and knowledge graph infrastructure is properly configured.

How does a knowledge graph improve contextual recall compared to standard vector search?

A knowledge graph improves contextual recall by structuring entity relationships in Neo4j alongside MongoDB RAG vector search. While vector search matches semantic similarity, the graph maps connections between entities discussed, enabling structured relationship queries.

Can I use this memory system to manage both ephemeral context and permanent project facts?

The memory system manages both ephemeral working notes and permanent facts. Core identity, project details, and system facts are stored persistently, while ephemeral working notes handle temporary contextual recall during active AI agent interactions.