Meerkats.ai MCP Server
Official# Meerkats.ai MCP Server
This is a Model Context Protocol (MCP) server for the Meerkats.ai web scraping service. It provides tools for web scraping, RSS feed processing, and Google search.
## Features
- **Web Scraping**: Scrape any URL and get the content as markdown
- **RSS Feed Processing**: Process RSS feeds and get the content
- **Google Search**: Search Google and get the results
## Installation
1. Clone this repository
2. Install dependencies:
```
npm install
```
3. Create a `.env` file based on `.env.example` and add your Meerkats.ai API key
4. Build the project:
```
npm run build
```
## Configuration
Add the MCP server to your MCP settings configuration file:
```json
{
"mcpServers": {
"meerkats": {
"command": "node",
"args": ["path/to/meerkats.ai/build/index.js"],
"env": {
"MEERKATS_API_KEY": "your_api_key_here",
"SCRAPPER_API_URL": "https://api.meerkats.ai",
"SCRAPPER_CALLBACK_API_URL": "https://api.meerkats.ai/callback"
}
}
}
}
```
## Available Tools
### scrape_url
Scrape a URL and return the content as markdown.
**Parameters:**
- `url` (required): URL to scrape
- `wait` (optional): Time to wait in seconds before scraping
- `query` (optional): Query to search for on the page
### rss_feed
Process an RSS feed and return the content.
**Parameters:**
- `url` (required): URL of the RSS feed
- `wait` (optional): Time to wait in seconds before processing
### google_search
Search Google and return the results.
**Parameters:**
- `query` (required): Query to search for on Google
## Example Usage
```javascript
// Example of using the scrape_url tool
const result = await use_mcp_tool({
server_name: "meerkats",
tool_name: "scrape_url",
arguments: {
url: "https://example.com",
wait: 2
}
});
// Example of using the google_search tool
const searchResult = await use_mcp_tool({
server_name: "meerkats",
tool_name: "google_search",
arguments: {
query: "meerkats.ai web scraping"
}
});
```
## License
MIT
TDQS
Scored across 10 tools
Several tools have overlapping purposes: google_places and google_map both return Google Maps data for location queries, and web_search and google_serp both provide search results. Additionally, guess_email and generate_support_emails overlap in guessing email patterns. This creates multiple ambiguous boundaries that an agent must navigate.
Tool names follow no consistent pattern: some start with verbs (verify_email, scrape_url), others with nouns (web_search, google_map), and some are mixed (generate_support_emails, check_domain_catch_all). There is no uniform verb_noun structure, making the naming style chaotic.
With 10 tools, the count is within the ideal range for a utility server. However, the presence of redundant or overlapping tools (e.g., google_places vs google_map) indicates a slight over-scoping, so it earns a 4 rather than a 5.
The email/domain verification tools cover verification, guessing, catch-all detection, and MX checks, while web tools cover search, scraping, and Google-specific queries. Minor gaps exist, such as missing reverse geocoding or additional domain checks, but core workflows are well covered.