# PhantomBuster MCP > PhantomBuster connects web data extraction and automation workflows directly to your AI agent. List all available scraping agents (Phantoms), launch new campaigns with specific arguments, check current usage limits, and retrieve the final extracted leads or structured data—all via natural conversation. ## Overview - **Category:** marketing-automation - **Price:** Free - **Tags:** automation, data-extraction, lead-gen, workflow-automation, browser-automation ## Description Web scraping used to be a clunky process involving dedicated dashboards and manual API calls. Now you can manage your entire lead generation pipeline using only chat commands. Instead of clicking through multiple tabs to see if a workflow is running, simply ask your agent about it. You can check the status of all your defined automation agents or launch an entirely new campaign on demand, supplying custom parameters in conversation. When scraping finishes, you don't have to hunt for the results; you just ask for the latest output and get the clean data ready for immediate use. Because Vinkius hosts this MCP, you access full control over all your web automation tools through one connection point. This means whether you need to audit a configuration or retrieve 10,000 leads from a past run, it's all available inside your agent. ## Tools ### abort_phantom Stops a web scraping agent that is currently running. ### get_phantombuster_account Retrieves your account's usage statistics and settings. ### get_phantom_setup Shows the specific configuration arguments used by a defined Phantom agent. ### get_phantom_output Fetches the final data and results from a completed scraping run. ### get_phantom Retrieves detailed information for one specific web scraping agent. ### get_workflow Gets details about an entire automated workflow setup. ### launch_phantom Starts a new Phantom execution with specified parameters. ### list_containers Lists all running and defined web scraping containers. ### list_phantoms Shows a list of every available Phantom agent you own. ### list_workflows Lists all the automated workflows configured in your account. ## Prompt Examples **Prompt:** ``` List all Phantoms in my account and show their current status. ``` **Response:** ``` I've retrieved 5 Phantoms: 'LinkedIn Search Export' (Success), 'Sales Navigator Search' (Error), 'Web Scraper' (Ready), and 2 others. Which one would you like to manage? ``` **Prompt:** ``` Launch the 'LinkedIn Search Export' Phantom with ID '12345'. ``` **Response:** ``` Phantom 'LinkedIn Search Export' (ID: 12345) has been launched successfully. I'll notify you once the output is ready. ``` **Prompt:** ``` Show me the extracted data from the last run of Phantom 12345. ``` **Response:** ``` I've fetched the results for Phantom 12345. It extracted 150 leads, including names, company titles, and LinkedIn profile URLs. Would you like a detailed breakdown? ``` ## Capabilities ### View and manage scraping agents List every defined web scraping tool (Phantom) and every automated workflow you have set up. ### Start or stop automation runs Launch a specific Phantom agent with custom arguments, or immediately halt a runaway process that's consuming credits. ### Retrieve extracted data results Fetch the final output and raw data collected from any completed scraping job for analysis. ### Audit automation configurations Review the precise setup arguments of a Phantom or workflow without running it, ensuring your logic is correct before launch. ### Check account usage and limits Get real-time information on your remaining execution time and overall account settings. ## Use Cases ### The Sales team needs a quick list of competitors. Instead of building a new workflow and hoping it works, the agent first uses get_phantom to check the details of an existing competitor profile Phantom. Once verified, they use launch_phantom to run it immediately with specific industry keywords. ### A developer needs to debug data extraction. The engineer runs a process that fails. They ask the agent to get_workflow details and check the corresponding Phantom configuration using get_phantom_setup, narrowing down whether the error is in the code or the input arguments. ### A marketing manager needs historical data. The manager asks for the output from a run last week. The agent fetches this specific result using get_phantom_output and provides 200 rows of structured lead data instantly, eliminating manual CSV downloads. ### A team needs to scale up scraping efforts. Before launching a massive campaign, the user first asks list_phantoms to ensure all necessary agents are present. They then check get_phantombuster_account to confirm they have enough capacity for the whole batch. ## Benefits - You gain immediate visibility into all automation assets. Instead of logging into a separate dashboard, simply ask the agent to list_phantoms or list_workflows to see every tool you own. - Stop manual intervention when things go wrong. If a scraping job gets stuck or runs too long, use abort_phantom to halt the process instantly and save credits. - Never lose data again. After a successful run, ask for the output using get_phantom_output; the agent pulls the clean results right into your conversation for you to review. - Verify complex logic before launching expensive campaigns. Use get_phantom_setup to audit exactly what parameters a Phantom needs without wasting execution time. - Track costs and capacity with confidence. Check your remaining credits and overall status using get_phantombuster_account, so you always know if you can afford the next scrape. ## How It Works The bottom line is you stop working in web dashboards and start directing your automation through conversation. 1. Subscribe to this MCP and provide your PhantomBuster API key credentials. 2. Tell your agent what you need, for example: 'List all Phantoms that scrape LinkedIn profiles.' 3. Your agent sends the request using the available tools, returning a list of agents or the requested data directly. ## Frequently Asked Questions **How do I list all my scraping agents using PhantomBuster MCP?** You ask your agent to list_phantoms. This tool shows you every defined web scraping agent (Phantom) in your account, letting you see what's available for use. **What if a scrape job is running too long? How do I stop it?** Use abort_phantom. You just tell the agent to abort_phantom and provide the Phantom name or ID, and it will immediately halt the running process. **Can I see what arguments a specific agent needs before launching it?** Yes, you use get_phantom_setup. This tool shows you exactly which configuration arguments are mandatory for that Phantom, saving you from errors and wasted credits. **Where do I find the data after running a PhantomBuster automation?** You retrieve it using get_phantom_output. Just tell your agent to fetch the output for a specific run ID or date, and the clean, extracted data appears right away. **Does this MCP help me track my web scraping costs?** Absolutely. Use get_phantombuster_account. This tool provides real-time account settings and usage information, so you always know your remaining execution time before launching a job.