Overview

McpDeepResearch is a lightweight MCP server that turns the familiar Google Scholar experience into a programmable API for academic discovery. By exposing three read‑only tools—, , and —it lets AI assistants quickly surface, retrieve, and present scholarly content without manual browsing. This capability is especially valuable for developers building research‑oriented chatbots, literature review assistants, or data‑collection pipelines that need reliable access to up‑to‑date papers.

The server solves the problem of navigating Google Scholar’s dynamic web interface and extracting clean text from research PDFs or HTML pages. It leverages a headless Chrome instance via the Chrome DevTools Protocol to perform searches and render pages, then parses the results into structured JSON or Markdown. Because all operations are read‑only and respect , developers can integrate it into institutional workflows that enforce strict data‑handling policies while still pulling the latest literature.

Key features include:

• Search with filters – accepts query strings, optional year ranges, and date sorting flags, returning a list of paper metadata (title, authors, publication venue, URL).
• Universal Markdown conversion – turns any public web page into clean, readable Markdown, stripping navigation bars and ads.
• Smart paper extraction – automatically detects the core content of a scholarly article (title, abstract, body, references) and removes extraneous elements, delivering a ready‑to‑read document.

Typical use cases involve academic assistants that can ask for the latest papers on a topic, pull them into the conversation, and summarize or analyze their content. For example, a research bot might search for “diffusion models 2023”, fetch the resulting PDFs, and then feed the Markdown to a summarization model. The server’s integration is seamless: any MCP‑compatible client can invoke these tools via simple JSON requests, whether over SSE or STDIO.

Unique advantages stem from its zero‑write policy and proxy support. Developers can run the server behind institutional proxies, ensuring compliance with network restrictions while still accessing external resources. Because it never writes files locally and only streams data back to the client, security concerns are minimized—making it a safe addition to sensitive research environments.