NYTimes MCP

"""NYT MCP Server - FastMCP-based server for New York Times API.""" from fastmcp import FastMCP from nytimes_mcp import resources, tools # Initialize FastMCP server mcp = FastMCP("nytimes-mcp", dependencies=["httpx", "pydantic", "python-dotenv"]) # Register all tools @mcp.tool() async def search_articles( query: str, sort: tools.SortOrder = "best", begin_date: str | None = None, end_date: str | None = None, page: int | None = None, ) -> dict: """ Search New York Times articles by query, date range, and other criteria. Args: query: Search query string sort: Sort order - "newest" or "oldest" (default: "newest") begin_date: Start date in YYYYMMDD format (optional) end_date: End date in YYYYMMDD format (optional) page: Page number for pagination, 0-indexed (optional) Returns: Formatted response with articles array containing headline, snippet, web_url, and pub_date """ return await tools.search_articles(query, sort, begin_date, end_date, page) @mcp.tool() async def get_latest_news( limit: int = 20, offset: int = 0, source: tools.NewsSource = "nyt", section: str = "all", ) -> dict: """ Get the latest news items from the NYT news wire (real-time news feed). Args: limit: Number of items to return (default: 20) offset: Pagination offset (default: 0) source: News source - "nyt" or "inyt" (default: "nyt") section: News section (default: "all"). e.g. "u.s." or "technology". Use the 'nyt://reference/sections' resource for available section names. Returns: Formatted response with news_items array containing title, abstract, url, section, subsection, published_date, and byline """ return await tools.get_news_wire(limit, offset, source, section) @mcp.tool() async def get_most_popular( popularity_type: tools.PopularityType = "viewed", time_period: tools.PopularityPeriod = "1", ) -> dict: """ Get the most popular New York Times articles. Args: type: Type of popularity - "viewed", "shared", or "emailed" (default: "viewed") time_period: Time period in days - "1", "7", or "30" (default: "1") Use the 'nyt://reference/popular-types' resource for available options. Returns: Formatted response with articles array containing title, abstract, url, and published_date """ return await tools.get_most_popular(popularity_type, time_period) @mcp.tool() async def get_archive(year: int | None = None, month: int | None = None) -> dict: """ Get New York Times articles from a specific month and year archive. Args: year: Year (default: current year) month: Month 1-12 (default: current month) Returns: Full NYT archive API response (unformatted) """ return await tools.get_archive(year, month) @mcp.tool() async def get_bestseller_list(list: str = "hardcover-fiction", offset: int = 0) -> dict: """ Get New York Times bestseller lists. Args: list: List name (e.g., "hardcover-fiction", "hardcover-nonfiction", "paperback-nonfiction") Default is "hardcover-fiction". Use the 'nyt://reference/bestseller-lists' resource for available list names. offset: Pagination offset (default: 0) Returns: Full NYT Books API response (unformatted) """ return await tools.get_bestseller_list(list, offset) # Register all resources @mcp.resource("nyt://reference/sections") def list_sections() -> dict: """Available sections for the get_top_stories tool.""" return resources.get_available_sections() @mcp.resource("nyt://reference/bestseller-lists") def list_bestseller_lists() -> dict: """Available bestseller list names for the get_bestseller_list tool.""" return resources.get_bestseller_lists() @mcp.resource("nyt://reference/api-limits") def list_api_limits() -> dict: """NYT API rate limits and usage information.""" return resources.get_api_limits() def main(): """Main entry point for the NYT MCP server.""" import asyncio # Cleanup handler for HTTP client async def cleanup(): await tools.cleanup_http_client() try: mcp.run() finally: # Run cleanup asyncio.run(cleanup()) if __name__ == "__main__": main()