agent-memory-mcp

Store and search persistent agent memory via a local Node.js MCP server.

Updated Mar 27, 2026
One-click install
npx skills add https://github.com/CenredJun/openclaw-claudecode-setup-kit --skill agent-memory-mcp-cenredjun
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-memory-mcp
Source: https://github.com/CenredJun/openclaw-claudecode-setup-kit/tree/main/skills/agent-memory-mcp
Command: npx skills add https://github.com/CenredJun/openclaw-claudecode-setup-kit --skill agent-memory-mcp-cenredjun

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill provides a persistent, searchable long-term memory and knowledge management layer so agents can retain architecture notes, design patterns, and decisions across sessions and projects for reliable retrieval.

Core Features & Use Cases

  • Persistent searchable memory: Indexes and stores architecture, patterns, and decision records for fast retrieval.
  • MCP server interface: Exposes memory_search, memory_write, memory_read, and memory_stats endpoints for programmatic agent access.
  • Project sync & dashboard: Automatically syncs with project documentation and includes a local dashboard to visualize memory usage and analytics.
  • Use Case: Capture architecture decisions during design meetings, then query historical patterns to guide future implementations or onboarding.

Quick Start

Clone the agentMemory repo into .agent/skills/agent-memory, install dependencies, compile, and start the MCP server for your project with the npm run start-server command providing a project_id and the absolute path to your workspace.

Frequently Asked Questions about agent-memory-mcp

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

FAQPage Schema
How do I add persistent searchable memory to an AI agent for retrieving architecture decisions?

You can achieve persistent searchable memory by running a local Node.js MCP server with memory_search, memory_write, memory_read, and memory_stats endpoints, allowing agents to index and retrieve architecture decisions and project documentation across sessions.

How do I sync project documentation to an agent knowledge base using MCP?

Syncing project documentation uses a local MCP server that automatically indexes workspace files into agent memory, exposing endpoints for programmatic read and write operations to keep architecture patterns updated.

Can I use an MCP memory server with Node.js to track design patterns across sessions?

Yes, the MCP memory server runs on Node.js v18+ and provides persistent storage to track design patterns across sessions, requiring you to clone the repository, install dependencies, and start the server with a project_id and workspace path.

What is the best way to provide long-term knowledge management for local development agents?

Long-term knowledge management for local development agents is best handled by a local MCP server that indexes decision records and architecture notes, including a dashboard to visualize memory usage and analytics within your development workspace.

Do I need Node.js v18 to run a local agent memory MCP server?

Yes, Node.js v18 or higher is required to run the local MCP server, which must be cloned into your workspace, compiled, and started using npm run start-server with a specific project_id and absolute path.