# Browse AI MCP

> Browse AI lets your agent scrape and monitor websites without writing code. You connect it via MCP to any client and give natural language commands to run web robots, check site changes, or pull structured data from pages you point to.

## Overview
- **Category:** data-management
- **Price:** Free
- **Tags:** data-extraction, no-code, web-monitoring, structured-data, automated-tasks, data-intelligence

## Description

Stop logging into dashboards just to see if a scraping job finished. This connection gives your AI agent full control over no-code web scraping and monitoring workflows using trained web robots. You can talk to the MCP and ask it to list all available scrapers, check past runs, or even start a new data collection task with custom parameters. Need to track changes on a specific page? Set up automated monitoring schedules, and your agent alerts you when something shifts. Furthermore, if you need visibility into your usage limits, you can programmatically check your account's credit balance. By connecting this MCP through the Vinkius catalog, your AI client acts like a dedicated data engineer, handling the entire process from task initiation to structured result delivery, all via natural conversation.

## Tools

### create_monitor
Sets up a new schedule to automatically check for changes on a specific website URL.

### remove_webhook
Deactivates an existing webhook configuration tied to a web scraping robot.

### get_user_profile
Retrieves basic account and user information for authentication purposes.

### get_robot_details
Gets specific, detailed configuration parameters for a single defined scraping robot.

### get_usage_quotas
Checks your current credit balance and automated data quota usage.

### get_run_status
Provides the real-time progress and status of a currently executing scraping task.

### list_bulk_operations
Fetches a comprehensive list of multiple completed or running data collection jobs.

### list_active_monitors
Retrieves a list of all currently active, scheduled website monitoring tasks.

### list_robots
Lists every configured web scraping robot available in your account for management.

### list_robot_history
List past runs

### list_robot_webhooks
Lists all active event notification configurations for your web scraping robots.

### trigger_robot_run
Initiates an immediate, new data collection task using a specified robot and custom URL.

## Prompt Examples

**Prompt:** 
```
List all available scraping robots in my account.
```

**Response:** 
```
I've retrieved your robots. You currently have 5 active scrapers, including 'Price Monitor' (ID: rob_123) and 'Lead Extractor'. Which one would you like to execute or inspect?
```

**Prompt:** 
```
Trigger robot 'rob_123' to scrape 'https://vinkius.com/pricing'.
```

**Response:** 
```
Scraping task triggered! I've initiated a new run for robot rob_123 with your custom URL. The task ID is 'task_456'. I'll monitor the progress and notify you once the results are ready.
```

**Prompt:** 
```
Check the status and results for task 'task_456'.
```

**Response:** 
```
Fetching results... Task task_456 has successfully 'Finished'. I've retrieved 12 data points including plan names and monthly prices. Would you like the full high-fidelity JSON export?
```

## Capabilities

### Manage scraping robots
List and get detailed information about every web scraping robot configured in your account.

### Execute data collection tasks
Start a new, custom scraping run on any specified URL at the moment you need the data.

### Monitor website changes
Set up and manage schedules to track specific areas of websites for automated change detection.

### Check task status and history
Get the current progress of a running job or retrieve detailed logs from past data collection runs.

### Configure event webhooks
Manage robot webhooks to receive instant notifications about specific events without needing to check a dashboard.

## Use Cases

### Tracking competitor pricing shifts
A market researcher needs to know if a competitor changed its 'Basic' plan price. Instead of manually checking the site every day, they use the MCP to `create_monitor` on the competitor's page. The agent handles the tracking and alerts them instantly when the price changes.

### Lead generation from public directories
A marketing manager needs a list of contact emails and phone numbers from 50 different local businesses. They use the MCP to `trigger_robot_run` with custom parameters, pulling structured data into their workspace instantly.

### Auditing compliance changes
A developer must ensure a critical legal page on a partner site hasn't changed since last week. They use the MCP to `list_active_monitors` and then manually check logs via `get_robot_history` to verify content integrity.

### Coordinating multi-source data builds
An analyst needs five different datasets from various sources. They use the MCP to first `list_robots` to confirm all scrapers are ready, then sequentially call `trigger_robot_run` for each source, coordinating a single comprehensive data build.

## Benefits

- Automate monitoring schedules by using the `create_monitor` tool. Instead of checking a dashboard daily, your agent tracks changes and alerts you when they happen on specific sites.
- Get immediate data without code. Use natural language to tell your agent to execute a task with `trigger_robot_run`, allowing you to start scraping immediately from your chat window.
- Track spending and quota limits by calling `get_usage_quotas`. This ensures you coordinate automated data collection efficiently and never overspend credits unexpectedly.
- Review massive amounts of historical data using the `get_robot_history` tool. You can instantly access past runs, saving time otherwise spent navigating through manual log pages.
- Maintain a perfectly coordinated pipeline by listing all scrapers with `list_robots` or checking active schedules with `list_active_monitors`, giving you full oversight of your data sources.

## How It Works

The bottom line is that you talk to your AI client like it talks to an API; the MCP translates those commands into scraping actions and delivers clean results back to the chat window.

1. Subscribe to this MCP and grab your Secret API Key from the Browse AI dashboard's account settings.
2. Pass the key to your preferred AI client, allowing it to interact with the web scraping tools via natural language commands.
3. Tell your agent what data you need—for instance, 'run robot X on URL Y'—and receive structured data or a progress update in response.

## Frequently Asked Questions

**How does the Browse AI MCP handle large amounts of scraped data?**
The agent retrieves structured results directly in your chat, allowing you to easily analyze and pass the clean output. You can also use `list_bulk_operations` to review multiple large task runs at once.

**Can I schedule monitoring using the Browse AI MCP?**
Yes, you use the `create_monitor` tool to set up automatic schedules. This ensures your agent tracks changes on a site and notifies you without requiring daily manual checks.

**What if I need to check my usage limits using Browse AI MCP?**
You can call `get_usage_quotas` within the conversation. This provides immediate visibility into your remaining credit balance, helping you plan future scraping efforts accurately.

**How do I see what robots are available with Browse AI MCP?**
Simply ask your agent to run `list_robots`. It will give you a complete list of every configured scraper in your account, letting you know exactly what data sources you have access to.

**Is the Browse AI MCP only for simple scraping?**
No. The MCP supports complex orchestration, allowing you to `list_robot_webhooks` and configure instant notifications that run outside of your direct interaction with the agent.