google_search
Perform targeted searches using a custom search engine by inputting specific terms or query arguments like site or date filters. Retrieve titles, links, and snippets for follow-up content extraction.
Instructions
Search the custom search engine using the search term. Regular query arguments can also be used, like appending site:reddit.com or after:2024-04-30. If available and/or requested, the links of the search results should be used in a follow-up request using a different tool to get the full content. Example: "claude.ai features site:reddit.com after:2024-04-30"
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| search_term | Yes |
Implementation Reference
- src/mcp_google_cse/server.py:54-82 (handler)The core implementation of the google_search tool. It builds a Google Custom Search service, executes the search query with the provided search_term and environment-configured parameters, cleans the snippets, formats results into TextContent objects with title, link, and snippet, and returns the list.async def google_search(search_term: str) -> List[TextContent]: """ Search the custom search engine using the search term. :param search_term: The search term to search for, equaling the q argument in Google's search. :return: Search results containing the title, link and snippet of the search result. """ service = build(os.getenv('SERVICE_NAME', 'customsearch'), "v1", developerKey=os.getenv('API_KEY')) response = service.cse().list( q=search_term, cx=os.getenv('ENGINE_ID'), cr=os.getenv('COUNTRY_REGION'), gl=os.getenv('GEOLOCATION', 'us'), lr=os.getenv('RESULT_LANGUAGE', 'lang_en'), num=os.getenv('RESULT_NUM', 10), fields='items(title,link,snippet)').execute() results = response['items'] __clean_up_snippets(results) text_contents = [] for result in results: text_contents.append( TextContent( type="text", text=f"Title: {result['title']}\nLink: {result['link']}\nSnippet: {result['snippet']}" ) ) return text_contents
- src/mcp_google_cse/server.py:16-35 (registration)Registers the 'google_search' tool with the MCP server via the list_tools handler, providing name, description, and input schema.@server.list_tools() async def handle_list_tools() -> list[Tool]: """ List available tools. Each tool specifies its arguments using JSON Schema validation. """ return [ Tool( name="google_search", description="Search the custom search engine using the search term. Regular query arguments can also be used, like appending site:reddit.com or after:2024-04-30. If available and/or requested, the links of the search results should be used in a follow-up request using a different tool to get the full content. Example: \"claude.ai features site:reddit.com after:2024-04-30\"", inputSchema={ "type": "object", "properties": { "search_term": {"type": "string"}, }, "required": ["search_term"], } ) ]
- src/mcp_google_cse/server.py:26-32 (schema)Defines the input JSON schema for the google_search tool, requiring a 'search_term' string property.inputSchema={ "type": "object", "properties": { "search_term": {"type": "string"}, }, "required": ["search_term"], }
- src/mcp_google_cse/server.py:85-93 (helper)Helper function called by the google_search handler to clean up snippets in search results by replacing non-breaking spaces (\xa0) with regular spaces and stripping trailing whitespace.def __clean_up_snippets(items: List[dict]) -> None: """ Remove non-breaking space and trailing whitespace from snippets. :param items: The search results that contain snippets that have to be cleaned up. :return: Nothing, the dict is mutable and updated directly. """ for item in items: item.update({k: v.replace('\xa0', ' ').strip() if k == 'snippet' else v for k, v in item.items()})