Provides tools to perform web searches using the Serper API, returning organic results, knowledge graph data, and related searches with support for advanced filters and operators.
Extracts content from web pages and transforms it into structured markdown, including support for metadata and JSON-LD extraction.
A TypeScript-based MCP server that provides web search and webpage scraping capabilities using the Serper API. This server integrates with Claude Desktop to enable powerful web search and content extraction features.
google_search - Perform web searches via Serper API
Rich search results including organic results, knowledge graph, "people also ask", and related searches
Supports region and language targeting
Optional parameters for location, pagination, time filters, and autocorrection
Supports advanced search operators:
site: Limit results to specific domain
filetype: Limit to specific file types (e.g., 'pdf', 'doc')
inurl: Search for pages with word in URL
intitle: Search for pages with word in title
related: Find similar websites
cache: View Google's cached version of a specific URL
before: Date before in YYYY-MM-DD format
after: Date after in YYYY-MM-DD format
exact: Exact phrase match
exclude: Terms to exclude from search results
or: Alternative terms (OR operator)
scrape - Extract content from web pages
Get plain text and optional markdown content
Includes JSON-LD and head metadata
Preserves document structure
Related MCP server: Serpex MCP
Node.js >= 18
Serper API key (set as SERPER_API_KEY environment variable)
Install dependencies:
npm installBuild the server:
npm run buildFor development with auto-rebuild:
npm run watchRun tests:
npm test # Run all tests
npm run test:watch # Run tests in watch mode
npm run test:coverage # Run tests with coverage
npm run test:integration # Run integration testsCreate a .env file in the root directory:
SERPER_API_KEY=your_api_key_hereSince MCP servers communicate over stdio, debugging can be challenging. We recommend using the MCP Inspector, which is available as a package script:
npm run inspectorThe Inspector will provide a URL to access debugging tools in your browser.
To install Serper Search and Scrape for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @marcopesani/mcp-server-serper --client claudeAdd the server config at:
MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json
{
"mcpServers": {
"serper-search": {
"command": "npx",
"args": ["-y", "serper-search-scrape-mcp-server"],
"env": {
"SERPER_API_KEY": "your_api_key_here"
}
}
}
}Open the Cline extension settings
Open "MCP Servers" tab
Click on "Configure MCP Servers"
Add the server config:
{
"mcpServers": {
"github.com/marcopesani/mcp-server-serper": {
"command": "npx",
"args": ["-y", "serper-search-scrape-mcp-server"],
"env": {
"SERPER_API_KEY": "your_api_key_here"
},
"disabled": false,
"autoApprove": ["google_search", "scrape"]
}
}
}Additional Cline configuration options:
disabled: Set to false to enable the server
autoApprove: List of tools that don't require explicit approval for each use
Open the Cursor settings
Open "Features" settings
In the "MCP Servers" section, click on "Add new MCP Server"
Choose a name, and select "command" as "Type"
In the "Command" field, enter the following:
env SERPER_API_KEY=your_api_key_here npx -y serper-search-scrape-mcp-serverYou can also run the server using Docker. First, build the image:
docker build -t mcp-server-serper .Then run the container with your Serper API key:
docker run -e SERPER_API_KEY=your_api_key_here mcp-server-serperAlternatively, if you have your environment variables in a .env file:
docker run --env-file .env mcp-server-serperFor development, you might want to mount your source code as a volume:
docker run -v $(pwd):/app --env-file .env mcp-server-serperNote: Make sure to replace your_api_key_here with your actual Serper API key.
Tool to perform web searches via Serper API and retrieve rich results. It is able to retrieve organic search results, people also ask, related searches, and knowledge graph.
| Name | Required | Description | Default |
|---|---|---|---|
| q | Yes | Search query string (e.g., 'artificial intelligence', 'climate change solutions') | |
| gl | Yes | Optional region code for search results in ISO 3166-1 alpha-2 format (e.g., 'us', 'gb', 'de') | |
| hl | Yes | Optional language code for search results in ISO 639-1 format (e.g., 'en', 'es', 'fr') | |
| location | No | Optional location for search results (e.g., 'SoHo, New York, United States', 'California, United States') | |
| num | No | Number of results to return (default: 10) | |
| tbs | No | Time-based search filter ('qdr:h' for past hour, 'qdr:d' for past day, 'qdr:w' for past week, 'qdr:m' for past month, 'qdr:y' for past year) | |
| page | No | Page number of results to return (default: 1) | |
| autocorrect | No | Whether to autocorrect spelling in query | |
| site | No | Limit results to specific domain (e.g., 'github.com', 'wikipedia.org') | |
| filetype | No | Limit to specific file types (e.g., 'pdf', 'doc', 'xls') | |
| inurl | No | Search for pages with word in URL (e.g., 'download', 'tutorial') | |
| intitle | No | Search for pages with word in title (e.g., 'review', 'how to') | |
| related | No | Find similar websites (e.g., 'github.com', 'stackoverflow.com') | |
| cache | No | View Google's cached version of a specific URL (e.g., 'example.com/page') | |
| before | No | Date before in YYYY-MM-DD format (e.g., '2024-01-01') | |
| after | No | Date after in YYYY-MM-DD format (e.g., '2023-01-01') | |
| exact | No | Exact phrase match (e.g., 'machine learning', 'quantum computing') | |
| exclude | No | Terms to exclude from search results as comma-separated string (e.g., 'spam,ads', 'beginner,basic') | |
| or | No | Alternative terms as comma-separated string (e.g., 'tutorial,guide,course', 'documentation,manual') |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden of behavioral disclosure. It mentions the tool 'retrieves rich results' but lacks critical details: it doesn't specify rate limits, authentication requirements, error handling, pagination behavior, or whether it's read-only or mutative. For a complex search tool with 19 parameters, this is a significant gap in transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is appropriately concise with two sentences that efficiently convey the core functionality. It's front-loaded with the main purpose and follows with specific result types. There's no wasted verbiage, though it could potentially benefit from more structured formatting given the tool's complexity.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (19 parameters), lack of annotations, and absence of an output schema, the description is incomplete. It doesn't explain what the return values look like, how results are structured, error conditions, or performance characteristics. For a search tool with rich parameter options, users need more context about what to expect from the operation.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema description coverage is 100%, with all 19 parameters well-documented in the input schema. The description doesn't add any parameter-specific information beyond what's already in the schema. According to guidelines, when schema coverage is high (>80%), the baseline score is 3 even without parameter details in the description.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool performs web searches via Serper API and retrieves rich results, specifying organic search results, people also ask, related searches, and knowledge graph. It uses specific verbs ('perform', 'retrieve') and identifies the resource (web searches). However, it doesn't explicitly differentiate from the sibling 'scrape' tool, which might have overlapping functionality.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus the sibling 'scrape' tool or other alternatives. It mentions the types of results retrieved but offers no context about appropriate use cases, prerequisites, or exclusions. This leaves the agent without clear decision-making criteria.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool to scrape a webpage and retrieve the text and, optionally, the markdown content. It will retrieve also the JSON-LD metadata and the head metadata.
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | The URL of the webpage to scrape. | |
| includeMarkdown | No | Whether to include markdown content. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of behavioral disclosure. It states what the tool retrieves (text, markdown, JSON-LD, head metadata) but lacks critical behavioral details such as rate limits, authentication needs, error handling, or whether it performs destructive actions (e.g., modifying data). For a web scraping tool with no annotation coverage, this is a significant gap in transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is concise and front-loaded in a single sentence, efficiently stating the core functionality without unnecessary details. Every part earns its place, though it could be slightly more structured (e.g., separating outputs). It avoids redundancy and is appropriately sized for the tool's complexity.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's moderate complexity (web scraping with two parameters), no annotations, and no output schema, the description is partially complete. It covers what the tool retrieves but misses behavioral aspects like rate limits or error handling. Without an output schema, it should ideally describe return values more explicitly, but it does list the types of content retrieved, providing some context.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema has 100% description coverage, clearly documenting both parameters ('url' and 'includeMarkdown'). The description adds no additional meaning beyond the schema—it doesn't explain parameter interactions, constraints, or usage examples. With high schema coverage, the baseline score of 3 is appropriate as the schema handles the heavy lifting.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'scrape a webpage and retrieve the text and, optionally, the markdown content. It will retrieve also the JSON-LD metadata and the head metadata.' This specifies the verb (scrape), resource (webpage), and outputs (text, markdown, JSON-LD, head metadata). However, it doesn't explicitly differentiate from the sibling tool 'google_search', which likely serves a different purpose (searching vs. scraping).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives. It doesn't mention the sibling tool 'google_search' or any other scraping-related tools, nor does it specify prerequisites, contexts, or exclusions for usage. This leaves the agent without clear direction on tool selection.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.2.0google_searchscrapeScored across 2 tools
The two tools have clearly distinct purposes: google_search performs web searches and retrieves search results, while scrape extracts content from a specific webpage. There is no overlap in functionality, making it easy for an agent to choose the correct tool based on the task.
Both tools follow a consistent snake_case naming convention with descriptive, action-oriented names (google_search and scrape). The naming pattern is uniform and predictable, enhancing readability and usability.
With only 2 tools, the server feels thin for a search and scrape domain, potentially lacking operations like advanced filtering, batch processing, or handling specific content types. While the tools cover core functionalities, the count is borderline minimal for the apparent scope.
The server provides basic search and scrape capabilities, but there are notable gaps such as no tools for managing search history, handling pagination in results, or extracting structured data beyond text/metadata. This may limit agents in complex workflows, though core operations are present.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Web search, fetch, extract, and research for AI agents. Markdown output + AI-synthesized answers.
Web search, URL content extraction to Markdown, site mapping, and recursive web crawler.
LLM-ready web search + instant answers + URL-to-clean-text fetch for agents and RAG.
Web search, scraping, Google Trends and data lookups. Paid per call in USDC on Base via x402.
