olostep/olostep-mcp-server

Overview

An MCP server that wraps Olostep's web scraping, crawling, batch extraction, and search infrastructure. It exposes these capabilities as tools for any MCP-compatible AI agent, enabling real-time web data access. The server is implemented in TypeScript and communicates via the Model Context Protocol.

Best for

Best for
Developers building MCP-compatible AI agents that need real-time web scraping and search capabilities

Use cases

• Scrape web pages for data extraction from an AI agent
• Crawl websites to collect structured content at scale
• Search the web in real time to retrieve current information

How to use

Install

npx -y @smithery/cli install @olostep/olostep-mcp-server --client claude

Tools exposed

• scrape_website
• search_web
• answers
• batch_scrape_urls
• create_crawl

Tested with

Cursor, Claude Desktop, Claude Code, Windsurf, VS Code, Metorial

Example client config

[object Object]

Notes

An MCP server that wraps Olostep’s web scraping, crawling, batch extraction, and search infrastructure. It exposes these capabilities as tools for any MCP-compatible AI agent, enabling real-time web data access. The server is implemented in TypeScript and communicates via the Model Context Protocol.

19 stars on GitHub. Last updated 2026-06-01. Licensed MIT.

Use cases

• Scrape web pages for data extraction from an AI agent
• Crawl websites to collect structured content at scale
• Search the web in real time to retrieve current information

Pros

• Provides a unified MCP interface for multiple web data operations
• Enables real-time web access without custom scraping code
• Supports batch extraction for large-scale data collection tasks

Cons

• Low community adoption (19 stars) may indicate limited testing or support
• Requires an Olostep API key and may incur usage costs
• Only works with MCP-compatible agents, limiting interoperability

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.