explaingit

liqlos/research-mcp-server

Analysis updated 2026-05-18

★ 0TypeScriptAudience · developerComplexity · 2/5Setup · easy

TLDR

A single MCP server that gives AI assistants like Claude or Cursor 23 pay-per-call research tools covering web search, social media, academic papers, and financial filings.

Mindmap

mindmap
  root((repo))
    What it does
      23 research tools
      One MCP endpoint
      Pay per successful call
    Tech stack
      TypeScript
      MCP protocol
      Apify hosting
    Use cases
      Web research
      Citation checking
      Trend detection
    Audience
      AI agent builders
      Researchers

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Give an AI coding assistant real web search and citation verification instead of guessed facts.

USE CASE 2

Chain tools together to research a topic: search, extract content, score reliability, and format a citation.

USE CASE 3

Monitor trending topics across Reddit, Hacker News, and news sources from inside an AI agent.

What is it built with?

TypeScriptMCPApify

How does it compare?

liqlos/research-mcp-server0xradioac7iv/tempfs52191314/web-agent-proxy-sdk
Stars000
LanguageTypeScriptTypeScriptTypeScript
Setup difficultyeasymoderatemoderate
Complexity2/53/54/5
Audiencedeveloperdeveloperdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you get it running?

Difficulty · easy Time to first run · 5min

Requires an Apify account token, usage is pay per successful call starting at $0.01.

In plain English

Research MCP Server gives an AI agent, such as Claude, Cursor, or ChatGPT, access to 23 different research tools through a single connection point called an MCP server. Instead of the AI making things up when it needs facts, it can call out to these tools to search the web, read academic papers, check social media, or pull financial filings, and get real data back. The tools are grouped by category. Web and content tools let it search the internet, pull clean text from any page, read Google News, get Wikipedia summaries, find archived versions of dead links, and score how reliable a source is. Social and discussion tools reach into Reddit, Hacker News, YouTube (including video transcripts), Substack newsletters, Bluesky, Telegram, Mastodon, and VKontakte, and can also detect what is trending across several of these at once. Academic tools search preprint servers like arXiv, bioRxiv, and medRxiv, search data repositories like Zenodo and Figshare, find papers that support or contradict a claim, verify citations against Crossref and OpenAlex, check a whole bibliography, and generate citations in formats like BibTeX or APA. A smaller set of specialized tools covers places, through OpenStreetMap, and company filings, through the SEC's EDGAR database. Each successful call costs a small fee, between one and three cents depending on the tool, and failed calls are free. You get a token, hook it up to your MCP client, and set which category of tools you want loaded so you are not wasting context on categories you will not use, with presets for web, social, academic, or specialized data. Setup is described as taking about two minutes: get the endpoint URL and a token, then configure your client, whether that is Claude Desktop, Cursor, ChatGPT, LangChain, or any other tool that speaks the MCP protocol. The README gives worked examples for research assistant workflows that chain several tools together, such as searching, extracting the article text, scoring its reliability, and formatting a citation, as well as separate chains for market intelligence, academic work, financial research, social listening, and location lookups.

Copy-paste prompts

Prompt 1
Show me how to connect the Research MCP Server to Claude Desktop and load only the web preset of tools.
Prompt 2
Write an MCP tool call chain using this server to search Reddit, extract an article, and verify its citations.
Prompt 3
Explain the difference between the web, social, academic, and data tool presets in this MCP server.

Frequently asked questions

What is research-mcp-server?

A single MCP server that gives AI assistants like Claude or Cursor 23 pay-per-call research tools covering web search, social media, academic papers, and financial filings.

What language is research-mcp-server written in?

Mainly TypeScript. The stack also includes TypeScript, MCP, Apify.

How hard is research-mcp-server to set up?

Setup difficulty is rated easy, with roughly 5min to a first successful run.

Who is research-mcp-server for?

Mainly developer.

Open on GitHub → Explain another repo

This repo across BitVibe Labs

Verify against the repo before relying on details.

Alternatives & similar repos