memory-management

Store patterns and retrieve them via HNSW vector search.

Updated Mar 5, 2026
One-click install
npx skills add https://github.com/fabri07/Vektor --skill memory-management-fabri07
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: memory-management
Source: https://github.com/fabri07/Vektor/tree/main/.agents/skills/memory-management
Command: npx skills add https://github.com/fabri07/Vektor --skill memory-management-fabri07

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @claude-flow/cli, and includes scripts (resource) and references (resource) components.

What problem does it solve?

AgentDB memory system with HNSW vector search. Provides fast pattern retrieval, persistent storage, and semantic lookup for learning and knowledge management.

Core Features & Use Cases

  • Store patterns and knowledge items for future reuse.
  • Semantic search to find similar past work.
  • Build a shared knowledge base across agents.

Quick Start

To start, initialize memory and store a sample pattern to verify setup.

Frequently Asked Questions about memory-management

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

FAQPage Schema
How do I add persistent memory to AI agents for semantic search?

Persistent memory for agents is enabled by storing patterns and knowledge items in a memory store, allowing agents to retrieve past work through HNSW vector search. This supports knowledge sharing and pattern reuse across multiple tasks.

What is HNSW vector search and how does it work for knowledge retrieval?

HNSW vector search enables fast semantic lookup by indexing stored patterns and knowledge items. It allows agents to find similar past work quickly, building a shared knowledge base for learning and knowledge management across tasks.

How do I build a shared knowledge base across multiple agents?

You can build a shared knowledge base across agents by storing patterns in a persistent memory store and using semantic search to retrieve them. This enables knowledge sharing and pattern reuse for future tasks among different agents.

Do I need @claude-flow/cli to use agent memory management?

Yes, @claude-flow/cli is required as a dependency to implement persistent memory for agents. It supports the scripts used for memory store initialization, HNSW vector indexing, and pattern storage.

Can I export and import stored patterns from the agent memory store?

Yes, the memory store supports optional export and import of stored patterns via scripts. This allows you to transfer knowledge items and semantic search data between different agent environments.

What are the limitations of using HNSW for agent memory persistence?

HNSW vector search provides fast pattern retrieval and semantic lookup but depends on dependency setup and script execution. It is designed for intermediate implementation depth and may require manual configuration for optimal indexing performance.