Server facts
websearch-mcp
Project
- Author
- mnhlt
- Repository
- github.com/mnhlt/WebSearch-MCP
- License
- Not specified
- Category
- Search & Web Data
Tech
- Runtime / language
- node
- Transports
- stdio
- Remote available
- No
- Authentication
- Not specified
Activity
- GitHub stars
- 42
- Last commit
- Data as of
Tags
- mcp
- search engine
- websearch
Overview
Web search inside an AI assistant is the focus of websearch-mcp, which connects MCP clients to a separate WebSearch Crawler API over stdio. Its web_search tool takes a required query and optional controls for result count, language, and region. Searches can be narrowed to selected domains, omit unwanted domains or terms, and target news, blogs, or all result types. The documented response example pairs titles and snippets with URLs, site names, and bylines, giving assistants search results to use in their responses.
Deployment involves both the MCP process and the crawler service that carries out searches. The setup documentation provides Docker Compose configurations for the crawler alongside FlareSolverr, including an Apple Silicon variation. Environment settings select the crawler API address and the default result limit. For integration and verification, the project documents MCP client configuration, direct crawler health and search checks, and an included interactive test client. It also shows how to request searches programmatically through an MCP client.
websearch-mcp setup notes
- websearch-mcp records stdio only; confirm a network transport upstream before planning remote access.
- websearch-mcp has no recorded license; confirm reuse terms in its repository.