# Zyte MCP for AI Agents AI Agent Connect

> Zyte MCP lets your AI agent scrape websites, automate browser actions, and pull search results without getting blocked. It handles the heavy lifting of proxy rotation and headless browser management, giving you clean data from any URL or search engine directly in your chat.

## Overview
- **Category:** data-analytics
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/undefined/ai-agent-connect
- **Tags:** web-extraction, serp-api, headless-browser, data-mining, proxy-management

## Description

The Zyte MCP lets you extract structured web data, perform browser automation, and fetch search engine results without the headache of getting blocked. You're often trying to get data from a site that's actively trying to kick you out. Usually, that means setting up proxies, managing headless browsers, and constantly updating your scraping logic. This Connector changes that. Instead of wrestling with infrastructure, you just tell your AI what to do. You can ask it to grab the latest headlines from a news site, navigate through a multi-step login, or pull organic search results for a specific keyword. It handles the how of web scraping, including the clicks, the scrolls, and the anti-bot bypasses, so you can focus on the what. When you connect this to your workflow via the Vinkius catalog, you're basically giving your agent a high-powered web browser that never gets tired or blocked. It's built for people who need to move fast. You don't have to worry about rotation logic or headless browser overhead. You just provide the URL and the goal. Whether you're pulling raw HTML for a project or navigating through complex menus to find a specific piece of information, it handles the heavy lifting. It's like having a data researcher who can bypass roadblocks and bring you back the exact numbers or text you need, formatted and ready to go.

## Tools

### get_stats
Retrieve your current API usage and success rates. Use this to keep an eye on your costs and consumption limits.

### extract_data
Pull web data from any target URL while handling browser rendering and anti-bot measures for you.

### search_data
Fetch organic search results from Google and other search domains for market research and SEO analysis.

## Prompt Examples

**Prompt:** 
```
Go to news.ycombinator.com and give me a summary of the top 5 stories.
```

**Response:** 
```
I've pulled the latest content from Hacker News. Here are the top 5 stories:

* **New AI Framework Released** - A new library for local LLM inference.
* **SpaceX Launch Update** - Details on the latest satellite deployment.
* **Tech Trends** - A deep dive into the current state of web3.
* **Coding Tips** - Best practices for Rust concurrency.
* **Market News** - Analysis of recent semiconductor price shifts.
```

**Prompt:** 
```
Search Google for 'best coffee machines' and show me the top 10 organic results.
```

**Response:** 
```
I've pulled the top 10 organic results for 'best coffee machines' from Google:

| Rank | Title | URL |
|---|---|---|
| 1 | **Wirecutter's Top Picks** | wirecutter.com/coffee |

| 2 | **Best Espresso Machines 2024** | techreview.com/espresso |
| 3 | **Coffee Roaster Reviews** | beanblog.com/reviews |
```

**Prompt:** 
```
How much have I spent on my Zyte API this month?
```

**Response:** 
```
I've checked your usage stats:

* **Total Requests:** 1,250
* **Success Rate:** 99.2%
* **Estimated Cost:** $5.40

Your success rate is currently well above the 95% target.
```

## Capabilities

### Extract raw HTML from any URL
Get the exact browser-rendered content from any target website.

### Perform clicks and scrolls
Navigate through complex web pages by automating actions like clicking buttons or scrolling for more data.

### Pull organic search results
Fetch structured results from Google and other search engines for market research.

### Take web page screenshots
Capture JPEG or PNG images of websites to verify content or layout.

### Route requests by country
View localized content by routing your requests through specific geographic locations.

### Monitor API usage
Check your current costs and success rates directly within the chat.

## Use Cases

### Competitive Intelligence
Ask your agent to check what your competitors are ranking for in the UK and give you the top 5 organic links.

### Dataset Building
Tell your agent to go to a retail site, click 'Next' until the end, and grab every product name into a list.

### Content Monitoring
Have your agent take a screenshot of a homepage to verify if a new banner or promotion is live.

### SEO Auditing
Compare Google search results for a specific keyword in five different countries to see regional differences.

## Benefits

- No more proxy management: Zyte handles the rotation and headers so your agent doesn't get blocked by websites.
- Real browser rendering: Use extract_data to get the exact HTML a human sees, not just the raw source code.
- Global reach: Route requests through different countries to see localized search results and content.
- Automated navigation: Use the browser automation tools to click buttons and scroll through pages automatically.
- Cost transparency: Track your spending and success rates directly through get_stats without leaving your chat.

## How It Works

The bottom line is you get clean, structured data from any site without managing the underlying scraping infrastructure.

1. Get your Zyte API keys ready from your dashboard.
2. Connect the Connector to your AI client via the Vinkius catalog.
3. Ask your agent to scrape a site, perform a search, or take a screenshot.

## Frequently Asked Questions

**Can the Zyte MCP bypass CAPTCHAs?**
Yes, this Connector handles anti-bot measures and proxy rotation automatically so your agent can get the data it needs without being blocked by common security hurdles.

**Does the Zyte MCP work on sites that require JavaScript?**
It does. The Connector uses browser rendering to ensure that the content your AI agent sees is the same as what a human sees in a real browser.

**Can I use the Zyte MCP to see search results from different countries?**
Yes, you can route your requests through specific countries to see localized search results and content from different geographic regions.

**How does the Zyte MCP handle browser actions like clicking?**
It performs complex sequences of actions like clicking buttons, scrolling through pages, and waiting for specific elements to load before extracting data.

**Is the Zyte MCP good for large-scale data mining?**
It's ideal for data mining because it manages the difficult infrastructure like proxy rotation and headless browser management for you.

**Can I see my API usage through the Zyte MCP?**
Yes, you can ask your agent to show your current usage, costs, and success rates directly in the chat by using the stats tool.

**Can I take a screenshot of a website using this server?**
Yes! Use the `extract_data` tool and set the `screenshot` parameter to true. The agent will return the visual state of the page as requested.

**How do I get organic search results from Google?**
Use the `search_data` tool, specifying 'google.com' as the domain and your keywords as the query. You can also set `maxResults` to control the list length.

**Is it possible to monitor my Zyte API spending?**
Yes, by using the `get_stats` tool. Note that this requires you to provide your `ZYTE_DASHBOARD_API_KEY` during setup.