BestReads MCP Server 📚🤔

What is BestReads MCP Server?

BestReads MCP Server is a remote MCP (Model Context Protocol) server built on Cloudflare Workers that delivers personalized book recommendations. It uses GitHub OAuth for authentication, Durable Objects to persist user preferences, and Workers AI to generate suggestions. It is intended for readers who want AI‑powered book discovery based on their tastes.

How to use BestReads MCP Server?

Connect your MCP client to either https://bestreads.dinas.workers.dev/sse (SSE) or https://bestreads.dinas.workers.dev/mcp (Streamable HTTP). After authenticating with GitHub, you can call the provided tools to manage your reading profile and receive recommendations.

Key features of BestReads MCP Server

• View reading history and profile with getProfile
• Add favorite genres via addGenre
• Add favorite authors with addFavoriteAuthor
• Track books you’ve read using addBookRead
• Mark disliked books and authors
• Reset all preferences with clearPreferences
• Get AI‑powered book recommendations

Use cases of BestReads MCP Server

• Building a personalized reading list based on your tastes
• Discovering new books while excluding disliked authors
• Tracking your reading history and preferences over time
• Integrating book recommendations into other MCP‑compatible tools

FAQ from BestReads MCP Server

What transports does BestReads MCP Server support?

It supports Server‑Sent Events (SSE) at /sse and Streamable HTTP at /mcp.

How are book recommendations generated?

Recommendations are generated using Cloudflare Workers AI, based on the user’s stored preferences.

Where is my reading data stored?

User preferences are persisted in Cloudflare Durable Objects, scoped to each authenticated user.

How do I authenticate with the server?

Authentication is handled via GitHub OAuth, using Cloudflare’s OAuth Provider library.

Can I deploy my own instance of BestReads MCP Server?

Yes. The repository includes a “Deploy to Cloudflare Workers” button and full instructions for cloning, configuring GitHub OAuth, and deploying with Wrangler.