# Cloud BOT MCP for AI Agents AI Agent Connect > Cloud BOT MCP lets you run automated browser tasks in the cloud. It handles web scraping, form filling, and navigating websites on autopilot. Instead of manually setting up scripts, you can tell your agent what to do and it handles the interaction with web pages for you. ## Overview - **Category:** productivity - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_oRL6ttBTfykVRqI7TcgoMA58YnKP0MfVILxJfBaK/ai-agent-connect - **Tags:** rpa, web-scraping, no-code, workflow-automation, cloud-execution, task-automation ## Description Imagine you have a repetitive task like pulling data from a messy website every morning or filling out dozens of identical forms. Usually, you'd need to write a complex scraper or sit there clicking through tabs all day. This Connector changes that by letting your agent talk directly to your Cloud BOT account. You can tell your agent to start a specific scraping job, check if it finished, or even grab files from your cloud storage. It basically turns your AI into a dedicated RPA engineer. You don't have to jump between different dashboards or deal with messy JSON files yourself. Since Vinkius hosts this Connector, you can easily connect it to your favorite client and start running these browser-based workflows with just a few words. It's about getting the work done without the manual overhead of managing the automation infrastructure. ## Tools ### cancel_job Stops a running automation job immediately if it's no longer needed. ### execute_bot Triggers a specific bot execution using custom JSON parameters for tasks like scraping or form filling. ### get_bot_details Pulls the specific configuration and metadata for a single bot in your account. ### get_job_status Checks the current progress and results of a specific automation job. ### list_bots Retrieves a full list of all available RPA bots within your Cloud BOT account. ### list_cloudbot_files Shows all files currently stored in your Cloud BOT storage area. ### list_jobs Displays a history of recent execution jobs and their results. ## Prompt Examples **Prompt:** ``` List all my browser robots. ``` **Response:** ``` I've found 3 bots in your Cloud BOT account: * **Price Scraper** (ID: bot_123) * **Lead Collector** * **Form Filler** Which one would you like to run? ``` **Prompt:** ``` Run the 'Price Scraper' on vinkius.com. ``` **Response:** ``` Starting the **Price Scraper** now. I've triggered the job for vinkius.com. **Job ID:** job_987 I'll keep an eye on it and let you know when it's done. ``` **Prompt:** ``` What's the status of job_987? ``` **Response:** ``` Checking... Job **job_987** finished successfully. * **Status:** Success * **Items Extracted:** 42 * **Storage:** Saved to `results.csv` in your cloud storage. Would you like me to list the files for you? ``` ## Capabilities ### Trigger browser bots Start a specific bot execution with custom inputs like URLs or JSON parameters. ### Monitor job progress Check the real-time status and final results of your automation jobs. ### Manage bot configurations View and manage all the browser robots available in your account. ### Access cloud storage List and manage files that your robots generate or use during tasks. ### Control running tasks Cancel or pause a running job immediately if it's no longer needed. ## Use Cases ### Competitor Price Tracking A growth marketer asks the agent to run a price scraper every morning and summarizes the results. ### Automated Lead Gen An ops lead uses the agent to trigger a bot that fills out a lead form on a partner site. ### Data Migration A dev uses the agent to list files in cloud storage to verify a batch of scraped records was saved. ### Status Reporting A manager asks the agent to list recent jobs to see if the nightly data sync succeeded. ## Benefits - Stop manual data entry. Use execute_bot to fill out complex web forms automatically. - Real-time monitoring. Use get_job_status to see exactly where your scraper is in the process. - Centralized bot management. Use list_bots to see your entire fleet of robots in one place. - Easy file access. Use list_cloudbot_files to grab the data your bots just finished extracting. - Instant job control. Use cancel_job to kill a runaway process without logging into a dashboard. ## How It Works The bottom line is you get a way to trigger and manage complex browser automations using only natural language. 1. Get your Access Token, Secret Key, and Public ID from the Cloud BOT dashboard. 2. Connect this Connector to your AI client using those credentials. 3. Ask your agent to run a bot, check a status, or list your files. ## Frequently Asked Questions **What is the Cloud BOT MCP?** It connects your AI agent to your Cloud BOT account so you can run browser tasks like scraping or form filling using just plain English. **Can I use it to scrape websites?** Yes, if you have a bot set up in Cloud BOT for a specific site, you can tell your agent to run it and get the data back. **Does it work with Claude or Cursor?** It works with any MCP-compatible client, including Claude, Cursor, and Windsurf, as long as you provide your API keys. **How do I see my bot results?** You can ask your agent to check the status of a specific job or list the files in your Cloud BOT storage to see what was saved. **Can I stop a bot if it's stuck?** Yes, your agent can use the cancel tool to stop any running job immediately if something goes wrong. **Do I need to write code to use this?** No, you just need to provide your Access Token, Secret Key, and Public ID from your dashboard. After that, you just talk to your agent. **How do I find my Cloud BOT API credentials?** Log in to your account, navigate to the **API Settings** section, and you will find your Access Token, Secret Key, and Public ID. **Can I pass custom parameters to a bot?** Yes! The `execute_bot` tool accepts a `params_json` string where you can provide a JSON object matching your bot's input configuration. **How do I check the results of an automation?** Use the `get_job_status` tool with a job ID to retrieve the execution status and any outputs or files generated by the bot.