T1nker-1220 Memories With Lessons
CreatedMCP-Mirror
Implements a knowledge graph memory server enabling Claude to remember user information and learn from past errors through a lesson system.
About
This server provides a basic implementation of persistent memory using a local knowledge graph. It enables the Claude AI model to remember information about users across chats and learn from past errors through a lesson system. It utilizes entities, relations, and observations to store information and provides an API with tools for managing these components, along with specific tools for creating, finding, updating, and recommending lessons based on error patterns and solutions. The server manages memory and lesson data in separate files and can be configured with Cursor MCP client and Claude Desktop.
Key Features
- Creates and manages entities, relations, and observations within a knowledge graph.
- 0 GitHub stars
- Offers lesson recommendations based on the current context.
- Provides tools for creating, finding similar, and updating the success of lessons.
- Manages memory and lesson data in separate files, splitting them automatically to maintain performance.
- Implements a lesson system for capturing knowledge about errors and their solutions.
Use Cases
- Creating a more context-aware and adaptive conversational AI experience.
- Improving AI model performance by allowing it to learn from past errors and successes.
- Personalizing chatbot interactions by storing user preferences and history.