Server facts
olostep-mcp-server
Project
- Author
- olostep
- Repository
- github.com/olostep/olostep-mcp-server
- License
- MIT
- Category
- Search & Web Data
Tech
- Runtime / language
- node
- Transports
- stdio
- Remote available
- No
- Authentication
- Not specified
Activity
- GitHub stars
- 24
- Last commit
- Data as of
Tags
- Content Extraction
- Web Scraping
- ai assistants
Overview
Web research and content collection take distinct paths in Olostep’s MCP integration. The search_web tool produces structured results through parsers, while answers uses web search to address factual questions with citations and an optional JSON output shape. For direct extraction, scrape_website supports HTML, Markdown, JSON, and plain text; get_webpage_content returns Markdown. Extraction options include JavaScript rendering, configurable delays, country-based requests, and specialized parsers. For link discovery rather than content retrieval, create_map lists site URLs with inclusion and exclusion filters, while get_website_urls finds links relevant to a query.
Collection workflows distinguish discovering pages from processing an existing URL list. create_crawl follows links from a starting page and launches a background job; get_crawl_results checks progress and retrieves the resulting content, with pagination and format selection. For a known collection of URLs, batch_scrape_urls starts a separate job whose output is retrieved through get_batch_results. The answers tool is intended for bounded factual questions, not dependable enumeration of changing page lists. Access requires an Olostep API key, with hosted access and local execution options documented for clients including Cursor, Claude Desktop, and VS Code.
olostep-mcp-server setup notes
- olostep-mcp-server records stdio only; confirm a network transport upstream before planning remote access.