# Octoparse MCP for AI Agents AI Agent Connect

> Octoparse lets you manage web scraping tasks through your AI client. It handles the heavy lifting of navigating complex websites, handling logins, and rendering JavaScript so you can focus on the data. Instead of jumping between tabs to check scraper statuses or manually downloading CSVs, you can just tell your agent to start a job, check the progress, and pull the latest results into your chat.

## Overview
- **Category:** industry-titans
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_HogQWNUeBfB97TxEMjMIKvpZMNIIERR4EpYDeK4q/ai-agent-connect
- **Tags:** data-extraction, no-code, web-automation, cloud-scraping, structured-data, data-ingestion

## Description

Managing web data usually feels like a chore of constant tab-switching. You start a scraper, wait for it to finish, then go hunt for the CSV file, only to find out it missed a few pages or hit a login wall. This Connector changes that by putting your account directly into your chat interface. You can talk to your agent to see what's currently running, pull specific data sets into your workspace, or check if a job finished successfully without ever opening a browser. It turns your scraping setup into a conversational data pipeline. By connecting this through the Vinkius catalog, you move from manual oversight to active orchestration. Your agent becomes the operator, handling the repetitive parts of data collection while you focus on the actual analysis.

## Tools

### get_new_data
Grab records that haven't been exported yet from a specific task. This helps you find fresh data quickly.

### get_task_data
Pull specific extracted data from a task using an offset for easy browsing. It lets you view data in chunks.

### get_task_status
See if a scraping job is running, completed, or stopped in real-time. Use this to monitor active jobs.

### list_task_groups
View all your organized scraping groups to keep your projects tidy. This helps you manage multiple projects.

### list_tasks
See all your scraping tasks, with the ability to filter by a specific group ID. This keeps your workspace clean.

### start_task
Trigger a cloud-based scraping job to begin gathering data from a website. Use this to initiate new scrapes.

### stop_task
Immediately halt a scraping task if it's behaving unexpectedly or finished early. Use this for manual control.

### update_data_status
Mark specific records as exported so you don't pull the same data twice. This ensures your data stays fresh.

## Prompt Examples

**Prompt:** 
```
Is my competitor pricing scraper finished?
```

**Response:** 
```
I've checked the status of your 'Competitor Pricing' task. It's currently **100% complete** and processed 1,240 records.

Would you like me to pull the latest results for you?
```

**Prompt:** 
```
Start the real estate leads scraper and tell me when it's running.
```

**Response:** 
```
Starting the 'Real Estate Leads' task now. I'll let you know as soon as it's active.

**Current Status:** Running
**Progress:** Beginning to crawl the first page.
```

**Prompt:** 
```
Show me all the tasks in my retail group.
```

**Response:** 
```
Here are the tasks currently in your Retail group:

| Task Name | Status | Last Update |
| :--- | :--- | :--- |
| Amazon Prices | Running | 2 mins ago |
| Walmart Stock | Stopped | 1 hour ago |
| eBay Trends | Completed | 5 hours ago |
```

## Capabilities

### Start and stop scraping jobs
Trigger cloud-based scraping tasks or halt them immediately from your chat.

### Check real-time scraper progress
See if a scraping job is running, completed, or stopped in real-time.

### Pull raw data into your chat
Retrieve specific extracted data or new records directly into your workspace.

### Organize and view scraping groups
View all your organized scraping groups to keep your projects tidy.

### Mark records as processed
Mark specific records as exported so you don't pull the same data twice.

## Use Cases

### Checking scraper progress
A researcher asks the agent to check the status of the Amazon Scraper and get a summary of progress using get_task_status.

### Pulling live pricing data
An analyst tells the agent to pull the last 50 records from a pricing task to review in the chat using get_task_data.

### Organizing multi-site scrapers
A dev asks to list all tasks in the Real Estate group to see which ones are active using list_tasks.

### Avoiding duplicate records
A user tells the agent to mark the latest batch of leads as exported to avoid duplicates using update_data_status.

## Benefits

- Stop manual CSV hunting by using get_task_data to pull results directly into your chat.
- Monitor scraper health in real-time with get_task_status instead of refreshing browser tabs.
- Orchestrate large-scale projects by using list_task_groups to keep your scrapers organized.
- Save time on repeated data pulls by using update_data_status to track what's already been processed.
- Rapidly trigger new data collection with start_task whenever you need fresh market info.

## How It Works

The bottom line is that you get a voice-controlled command center for your web scraping infrastructure.

1. Connect your Octoparse OpenAPI Access Token from your profile settings.
2. Link the connection to your favorite MCP client like Claude or Cursor.
3. Ask your agent to list your tasks or start a specific scrape.

## Frequently Asked Questions

**How can I use the Octoparse MCP to manage my scrapers?**
You can use it to start, stop, and monitor your Octoparse tasks through natural conversation.

**Can the Octoparse MCP pull data directly into my chat?**
Yes, you can ask your agent to retrieve specific data sets or new records directly from your tasks.

**Does Octoparse MCP work for monitoring competitor prices?**
It's great for that. You can set up a scraper for pricing and ask your agent to check its status and fetch updates regularly.

**How do I keep my Octoparse tasks organized with this Connector?**
You can use the list commands to view your task groups and specific tasks, making it easy to stay organized.

**Can I stop a running scraper using the Octoparse MCP?**
Yes, you can tell your agent to stop any active scraping task immediately if you need to.

**Does the Octoparse MCP help with avoiding duplicate data?**
Yes, by using the status update tool, you can mark records as exported so your agent knows what's already been processed.

**Can my AI automatically find the latest extracted data for a specific task?**
Yes! Use the `get_not_exported_data` tool with the Task ID. Your agent will respond with complete metadata for the newest records that haven't been marked as exported yet in seconds.

**How do I find my Octoparse OpenAPI Access Token?**
Log in to Octoparse, navigate to the **OpenAPI** section in your profile or developer portal, and follow the instructions to generate a Bearer token using your account credentials.

**Can I start a scraper via the AI?**
Absolutely. Use the `start_task` tool with your Task ID. The AI will command Octoparse to begin the extraction in the cloud immediately.