Research MCP Server — 10 Tools for AI Agents

Pricing

Rating

Developer

Actor stats

Categories

Share

One MCP endpoint that gives any AI agent ten fast research tools. Point your MCP client (Claude Desktop, Claude Code, ChatGPT, Cursor, or any Model Context Protocol app) at this Actor's URL and your agent can instantly search papers, code, news, forums, crypto markets, and more — no per-tool setup, no glue code.

This is a hosted MCP server, not a scraper you run once. Connect once and the ten tools show up in your agent automatically via MCP tool discovery.

The 10 tools

Each tool returns clean, structured JSON — ready for an agent to reason over.

How to connect

This Actor runs as a Standby MCP server over HTTP (Streamable HTTP transport). Use its Standby URL as the MCP server endpoint, authenticated with your Apify token:

https://dami-studio--research-mcp-server.apify.actor/
Authorization: Bearer <YOUR_APIFY_TOKEN>

Claude Desktop / Cursor (mcp config):

{
"mcpServers":{
"research":{
"url":"https://dami-studio--research-mcp-server.apify.actor/",
"headers":{"Authorization":"Bearer <YOUR_APIFY_TOKEN>"}
}
}
}

Then ask your agent things like "search arXiv for recent papers on retrieval-augmented generation" or "what's trending on Hacker News about Rust?" — it picks the right tool automatically.

First call may be slow. The server and each tool warm up on first use (Standby cold start); subsequent calls are fast. If a first call times out, just retry.

Why this exists

AI agents are far more useful when they can pull live external data. Wiring up ten separate research APIs (each with its own auth, rate limits, and response shape) is tedious. This server collapses all of that into one URL and one token — ten consistent, structured research tools your agent can call out of the box.

Great for: research assistants, market/competitive monitoring agents, coding agents that need to check GitHub/Stack Overflow/HN, and any Claude/ChatGPT workflow that benefits from live data.

Notes

• Built on the official Model Context Protocol SDK (McpServer + Streamable HTTP, stateless per request).
• Each tool is a thin, low-latency proxy to a dedicated, hardened source Actor — so results are clean and the toolset is easy to extend.
• Public sources only; no login or scraping of private data.