OneSearch MCP Server

MCPOfficialOpen SourceMIT51.5

by yokingma • Search & Retrieval

An open-source MCP server that provides web search, crawling, scraping, and content extraction via multiple search engines and Firecrawl.

Example Use Cases

1
Up-to-date web search results aggregated from multiple providers (SearxNG, DuckDuckGo, Bing, Tavily) for information retrieval tasks.
2
Crawl and scrape full web pages or sites (using Firecrawl or puppeteer-based local browser scraping) to extract structured content.
3
A self-hostable, configurable MCP service to integrate web search and scraping into agent workflows without depending solely on commercial APIs.

Description

OneSearch MCP Server integrates with SearxNG, Tavily, DuckDuckGo, Bing and local browser scraping (via puppeteer) to perform web search, crawl, scrape and extract content. It supports self-hosted components (SearxNG, Firecrawl) and can be run locally or connected to cloud APIs, exposing tools like one_search, one_scrape and one_map. The project is designed for agents and MCP-compatible clients that need up-to-date web data and site extraction without relying solely on paid third-party services. It also documents environment variables and deployment options for running on platforms such as Cursor and Windsurf.

Capabilities(4 total)

Tools (4)

one_search

Search and retrieve content from web pages. Returns SERP results by default (url, title, description).

query:string*limit:numberlanguage:stringcategories:stringtimeRange:string

one_scrape

Scrape a single webpage with advanced options for content extraction. Supports various formats including markdown, HTML, and screenshots. Can execute custom actions like clicking or scrolling before scraping.

url:string*formats:arrayonlyMainContent:booleanincludeTags:arrayexcludeTags:arraywaitFor:numbertimeout:numberactions:arrayextract:objectmobile:booleanskipTlsVerification:booleanremoveBase64Images:booleanlocation:object

one_map

Discover URLs from a starting point. Can use both sitemap.xml and HTML link discovery.

url:string*search:stringignoreSitemap:booleansitemapOnly:booleanincludeSubdomains:booleanlimit:number

one_extract

Extract structured information from web pages using LLM. Supports both cloud AI and self-hosted LLM extraction.

urls:array*prompt:stringsystemPrompt:stringschema:objectallowExternalLinks:booleanenableWebSearch:booleanincludeSubdomains:boolean

Quick Actions

View on GitHub

Security

Scanned 1 month(s) ago

Risk Level

MINIMAL

Read-only data retrieval, no side effects

Trust Score

C58/100

12/17 checks passed

Scores are informational only and provided “as is” without warranty. AgentHotspot assumes no liability for actions taken based on these ratings.

Quick Stats

Service TypeMCP

Pricing ModelFree

Capabilities4 Tools / 0 Prompts / 0 Resources

Owneryokingma

CategorySearch & Retrieval

DependenciesStandalone

Set Your Username