MCP server for Serper API integration with web search, news, images, videos, maps, reviews, and FireCrawl web scraping capabilities
Server defines 8 tools with complete input schemas and descriptions. Naming follows verb_noun convention (serper_*, firecrawl_*). Descriptions are detailed (100-250 chars) and explain WHEN to use each tool. However, output schemas are entirely undocumented, LLMs cannot predict response structure for chaining. No error handling guidance. Parameters lack validation constraints (e.g., num should enforce 1-20 range). Missing idempotency hints and recovery guidance. Composition is sound (each tool has one responsibility), but response field naming and pagination strategy are not documented.
Scrapes web pages using the FireCrawl API to extract content in a structured format. Ideal for retrieving the full content of articles, documentation, or any web page found in search results. Returns content in markdown format for easy readability. Use this tool when you need detailed information from a specific website.
Gets the current GPS coordinates using browser geolocation or fallback IP-based location services. Provides latitude and longitude that can be used for location-aware searches and services. No input parameters required - simply returns the current detected location. Note: Accuracy may vary depending on the device and available location services.
Searches for images using the Serper.dev Images API. Ideal for finding visuals, diagrams, photos, logos, and other graphical content. Returns detailed information including: - Image URLs and thumbnails - Image dimensions - Source websites Use this when you need visual content related to a search query.
Searches for places, businesses, and points of interest using the Serper.dev Maps API. Ideal for finding nearby locations, addresses, business information, and more. Returns detailed place information including: - Business names and addresses - Geographic coordinates (latitude/longitude) - Ratings and reviews - Business types and categories - Contact information when available Use this when you need to find specific places or search for places near a location.
Output schemas completely undocumented. LLMs cannot predict response structure, field names, or types. Breaks tool chaining, agent cannot extract IDs or references for downstream calls.
Numeric parameters (num: 1-20) lack min/max constraints in schema. LLMs may pass invalid values (0, 100, -1). No validation guidance in descriptions.
No error handling guidance. Tool descriptions do not explain what to do if API fails, rate limit hit, or invalid input provided. Agents cannot self-correct.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 57 | 2026-07-28+ | v2 |
Searches for news articles using the Serper.dev News API. Best for recent news articles, press releases, and timely content. Returns detailed information including: - Article titles and snippets - Source names and publication dates - Links to articles Use this when you need up-to-date news information on specific topics.
Retrieves detailed user reviews for a specific place or business from the Serper.dev Reviews API. This tool requires identification parameters from a previous maps search result. Returns detailed review information including: - Rating and date - Review text and likes - Reviewer information (name, profile) - Sorting options for different review views Use this to research user experiences, sentiment, and feedback about places.
Searches for videos using the Serper.dev Videos API. Great for finding video content from platforms like YouTube and other video hosting sites. Returns information including: - Video titles and descriptions - Source links - Channel information when available Use this when you need video content related to a search query.
Performs a web search using the Serper.dev API, ideal for general queries, news, articles, and online content. Use this for broad information gathering, recent events, or when you need diverse web sources. Supports location-based search and language/region customization. Maximum 20 results per request.
serper_reviews_search requires placeId but description does not explain how to obtain it (from serper_maps_search). Undocumented dependency forces agent to guess or fail.
No pagination or result limit guidance. Descriptions state 'Maximum 20 results' but do not explain how to fetch next page, whether results are sorted, or what happens if query matches >20 items.