DuckDuckGo & Felo AI Search MCP

The DuckDuckGo & Felo AI Search MCP is a privacy‑first, high‑performance search server that plugs directly into any Model Context Protocol (MCP)–compatible AI assistant. By scraping DuckDuckGo’s public HTML pages and augmenting results with Felo AI’s generative search engine, it delivers richer, more up‑to‑date information than typical API‑based tools. Developers can therefore empower assistants with real‑world web knowledge without exposing user data to third‑party services or requiring costly API keys.

At its core, the server exposes a set of intuitive tools that cover the full search workflow: a web‑search tool for standard keyword queries, an felo-search tool that leverages AI to interpret ambiguous prompts, and utilities for extracting URL content and metadata. Each tool accepts simple parameters—query strings, page numbers, or result limits—and returns structured JSON that can be seamlessly incorporated into conversational contexts. The server’s caching layer and rotating user agents protect against rate limits while maintaining low latency, making it suitable for both interactive chats and batch data gathering.

Key capabilities include:

• Privacy‑friendly scraping of DuckDuckGo results, avoiding the restrictions and privacy concerns of official APIs.
• AI‑enhanced search via Felo AI, allowing assistants to answer nuanced questions with contextually relevant snippets.
• Smart content extraction that filters out boilerplate and retrieves the most useful text, images, titles, and descriptions from a URL.
• Performance optimizations such as in‑memory caching and configurable rate limiting, ensuring consistent response times even under load.
• Zero‑configuration operation—no API keys or credentials are required, and the server can be launched with a single command.

Real‑world use cases span from building knowledge‑base bots that fetch the latest news, to powering research assistants that pull technical documentation directly from the web. Because it follows the MCP standard, any assistant that supports MCP can simply declare a new server in its configuration and start issuing search requests without custom adapters. This plug‑and‑play nature accelerates prototyping, reduces maintenance overhead, and keeps data handling within the user’s own infrastructure.

In summary, the DuckDuckGo & Felo AI Search MCP gives developers a secure, efficient, and fully integrated search capability for AI assistants. It bridges the gap between static knowledge bases and dynamic web content, enabling conversational agents to provide timely, accurate, and privacy‑respecting answers in a single, cohesive workflow.