# Browse AI MCP for AI Agents MCP

> Browse AI automates web data extraction by allowing your AI agent to run complex scraping robots and monitor website changes. You can trigger single-page extractions, perform bulk operations across dozens of URLs, or check for updates on competitor websites—all through natural conversation.

## Overview
- **Category:** data-management
- **Price:** Free
- **Tags:** web-extraction, no-code, data-monitoring, automation-robots, structured-data, web-scraping

## Description

Need structured data from a webpage but don't want to write Python scripts? This MCP lets your AI client take full control of web scraping tasks. Instead of manually navigating dashboards and copy-pasting metrics, you simply tell your agent what data you need. It executes the necessary extraction robots on target websites and delivers the final result—structured JSON data—right into your conversation context.

This isn't just a simple fetch; it manages everything from triggering multi-URL runs to checking if specific elements on a page have changed since last week. If you're using Vinkius, you can connect this powerful web intelligence source alongside thousands of others in the catalog, making your agent a true data pipeline.

## Tools

### run_bulk_task
Runs an extraction robot across multiple URLs concurrently to collect data from many sources at once.

### download_bulk_data
Gathers and returns all extracted results into a single unified JSON array after a multi-task run completes.

### get_bulk_task
Checks the execution status of a large, multi-URL bulk scraping operation.

### get_robot
Retrieves the detailed setup and configuration for a specific web scraping robot.

### get_task_data
Downloads the final, structured JSON data payload from a successful web scraping task.

### get_task
Checks the current status of any specific data extraction task run through Browse AI.

### list_credits
Retrieves your current usage count and quota limits for API credits.

### list_monitors
Lists all currently active web monitoring robots that check for page changes on a schedule.

### list_robots
Lists all your active extraction and monitoring robots configured in Browse AI.

### run_robot
Triggers a configured robot to extract data from a specified target URL, returning a task ID.

## Prompt Examples

**Prompt:** 
```
Can I pull all product specs from three different vendor sites at once?
```

**Response:** 
```
You got it. I'll run the bulk extraction now for those URLs. 

*Status:* Running 3/3 tasks...
*Progress:* 65% complete.

I've initiated the `run_bulk_task`. Give me a minute, and once done, I'll aggregate all the data into one giant JSON array ready for you to download.
```

**Prompt:** 
```
Which of my robots are set up to monitor changes?
```

**Response:** 
```
I checked your monitoring list. You have two active web monitors:

* **Competitor Price Tracker**: Checks the main pricing page every 4 hours.
* **Product Feature Watcher**: Monitors the 'Specs' tab for key feature changes daily.

The Competitor Price Tracker is currently reporting no change since its last run.
```

**Prompt:** 
```
Give me all the extracted data from task t-56d3f after it finished.
```

**Response:** 
```
Task `t-56d3f` completed successfully. Here are the top 3 items found:

| Product | Price (USD) | Status |
| :--- | :--- | :--- |
| Widget Pro | 199.00 | Available |
| Basic Tool | 49.50 | Low Stock |
| Elite Model | 799.00 | Sold Out |

I have the full, detailed JSON array stored in context if you want me to filter this list by 'Low Stock' items.
```

## Capabilities

### Execute Web Scraping Workflows
Trigger specific robots to run tasks on target URLs or initiate multi-URL concurrent extractions.

### Identify and Check Robots
List all your trained extraction workflows, view their configuration details, or check for the status of running jobs.

### Retrieve Extracted Data and Status
Check the progress of a specific scraping task or instantly download the final extracted JSON data from any successful run.

### Monitor Web Changes
List all active web monitoring robots and check if target pages have changed since their last scheduled scan.

### Manage API Usage
Retrieve current credit usage against your monthly plan limits for planning bulk operations.

## Use Cases

### Monitoring Competitor Pricing Changes
A marketing manager needs to know if a competitor changed their pricing structure. Instead of manually visiting the product page daily, they connect this MCP and use `list_monitors` to set up an automated check. The AI agent then alerts them immediately when prices deviate from the stored baseline.

### Analyzing Market Trends Across Multiple Sites
A data analyst needs a dataset of product specifications from ten different vendor websites. They use `run_bulk_task` to send all ten URLs to the robot simultaneously and then retrieve the unified results using `download_bulk_data`, which saves hours of individual scraping.

### Verifying Large Data Sources for Research
A researcher needs confirmation that a large data source is stable. They check the robot configuration using `get_robot` to verify the fields, then use `run_robot` to test it on a new URL, ensuring the output structure matches their expectations.

## Benefits

- Get real-time data results instantly. After running a job using `get_task_data`, you get the structured JSON directly in your chat, skipping manual downloads.
- Stop wasting time on single pages. Use `run_bulk_task` to run an extraction robot across dozens of URLs simultaneously and download all results with `download_bulk_data`.
- Stay ahead of competitors without constant checking. Use the monitoring tools to watch for specific changes, getting notified only when necessary.
- Know your limits upfront. The `list_credits` tool lets you check API usage before launching a massive data pull, saving you from unexpected overages.
- Simplify complex workflows by grouping related actions. You can list all available workflows with `list_robots` and then trigger them without needing to remember every function name.

## How It Works

The bottom line is: your AI client handles all the complex steps of scraping, monitoring, and data retrieval so you don't have to interact with a separate dashboard.

1. First, connect this MCP and provide your Browse AI Secret API Key to your agent.
2. Next, tell your agent which robot you want to run or what data you need. Your agent will use the necessary tools to initiate the extraction task on the target URL(s).
3. Finally, once the job completes, your agent provides the structured JSON payload directly, allowing you to analyze and act on the live web intelligence.

## Frequently Asked Questions

**Can my AI start a data extraction and let me know when it finishes?**
Yes! You can ask your agent to `run_robot` on a certain URL. The agent receives the newly created task ID and can immediately invoke `get_task` sequentially to poll for completion. Once successful, it calls `get_task_data` to dump the JSON right into your conversation without breaking workflow.

**How do bulk operations work natively through the agent?**
When you have a trained robot, you can feed the `run_bulk_task` command an array of different URLs. Browse AI's backend splits them virtually into independent task instances. Your agent tracks the bulk run via `get_bulk_task` and eventually pulls the mega JSON merge using `download_bulk_data`.

**Is there a risk of my agent suddenly spending all my API credits?**
You have complete visibility—the agent can query `list_credits` at any moment to inform you precisely of your available balance and monthly limit. Also, since tools execute your natural directives, your AI agent only triggers extractions you specifically prompt for.