The Pensieve MCP Server is a TypeScript-based implementation of a Retrieval-Augmented Generation (RAG) knowledge management system. It allows users to store, organize, and query knowledge using natural language, powered by Large Language Models (LLMs). The server provides a structured way to access and analyze stored knowledge, making it easier to spot patterns and connections.
memory://
URIs.ask_pensieve
: A natural language query tool.To get started, install the dependencies:
npm install
Build the server using:
npm run build
For development with auto-rebuild, use:
npm run watch
Store your knowledge as Markdown files in the root directory:
1. Use clear categories in filenames (e.g., skills-javascript.md
).
2. One topic per file.
3. Include relevant metadata (dates, sources, etc.).
4. Use clear, well-structured content.
To use The Pensieve MCP Server with Claude Desktop, add the server configuration:
On MacOS:
~/Library/Application Support/Claude/claude_desktop_config.json
On Windows:
%APPDATA%/Claude/claude_desktop_config.json
{
"mcpServers": {
"The Pensieve": {
"command": "/path/to/The Pensieve/build/index.js"
}
}
}
Debugging MCP servers can be challenging due to stdio communication. Use the MCP Inspector for debugging:
npm run inspector
The Inspector provides a URL to access debugging tools in your browser.
The Pensieve MCP Server is designed to help users manage and analyze their knowledge effectively. By storing thoughts and information in a structured format, users can easily retrieve and examine them at their leisure.
No releases published.
No packages published.