Memories With Lessons
What is the Knowledge Graph Memory Server?
This server provides Claude with long-term memory capabilities by storing information in a structured knowledge graph format. It remembers user details, preferences, and past interactions, while also capturing error patterns and solutions for continuous learning.How does it work?
The server stores information as interconnected entities (people, organizations, events) with observations (facts) and relationships. Special 'lesson' entities capture error patterns and verified solutions for future reference.When should I use it?
Ideal for: 1) Personalized AI assistants that remember user details 2) Technical support bots that learn from past solutions 3) Any application requiring persistent memory across conversations.Key Features
Entity SystemStore information about people, organizations, and events as interconnected nodes with attributes and relationships.
Error LearningCapture error patterns and solutions as special 'lesson' entities that improve over time with success rate tracking.
Persistent MemoryMaintain information across different conversations through local JSON storage that persists between sessions.
Cursor MCP IntegrationSeamless integration with Cursor IDE through Model Context Protocol for developer assistance scenarios.
Pros and Cons
Advantages
Remembers user preferences and details across conversations
Learns from mistakes by capturing error patterns and solutions
Lightweight local storage avoids cloud dependencies
Structured knowledge graph enables complex relationship modeling
Limitations
Local storage limits sharing across devices
Manual setup required for Cursor integration
Basic search capabilities (no advanced semantic search)
File-based storage may have performance limits at scale
Setup Guide
Installation
Install via Docker or NPX depending on your preference.
Cursor Configuration
Add the server configuration to your claude_desktop_config.json file.
System Prompt
Configure Claude's behavior with a prompt template that specifies when and how to use memory.
Memory Management
Use the provided API tools to create, read, update and delete memory entries as needed.
Example Scenarios
Personal AssistantClaude remembers user preferences and personal details across conversations
Technical SupportLearn from and recall solutions to common technical errors
Project ManagementTrack team members and their relationships to projects
Frequently Asked Questions
Where is the memory data stored?
How do I clear the memory?
Can I use this with multiple users?
What happens if the JSON files get too large?
Additional Resources
GitHub Repository
Original source code and documentation
Model Context Protocol
About the MCP standard this implements
Knowledge Graphs Explained
Background on knowledge graph technology
精選MCP服務推薦

Baidu Map
已認證
百度地圖MCP Server是國內首個兼容MCP協議的地圖服務,提供地理編碼、路線規劃等10個標準化API接口,支持Python和Typescript快速接入,賦能智能體實現地圖相關功能。
Python
717
4.5分

Markdownify MCP
Markdownify是一個多功能文件轉換服務,支持將PDF、圖片、音頻等多種格式及網頁內容轉換為Markdown格式。
TypeScript
1.7K
5分

Firecrawl MCP Server
Firecrawl MCP Server是一個集成Firecrawl網頁抓取能力的模型上下文協議服務器,提供豐富的網頁抓取、搜索和內容提取功能。
TypeScript
3.9K
5分

Sequential Thinking MCP Server
一個基於MCP協議的結構化思維服務器,通過定義思考階段幫助分解複雜問題並生成總結
Python
285
4.5分

Magic MCP
Magic Component Platform (MCP) 是一個AI驅動的UI組件生成工具,通過自然語言描述幫助開發者快速創建現代化UI組件,支持多種IDE集成。
JavaScript
1.7K
5分

Context7
Context7 MCP是一個為AI編程助手提供即時、版本特定文檔和代碼示例的服務,通過Model Context Protocol直接集成到提示中,解決LLM使用過時信息的問題。
TypeScript
5.2K
4.7分

Edgeone Pages MCP Server
EdgeOne Pages MCP是一個通過MCP協議快速部署HTML內容到EdgeOne Pages並獲取公開URL的服務
TypeScript
260
4.8分

Notion Api MCP
已認證
一個基於Python的MCP服務器,通過Notion API提供高級待辦事項管理和內容組織功能,實現AI模型與Notion的無縫集成。
Python
130
4.5分