PhantomBuster MCP. Manage and extract lead data from web pages.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
PhantomBuster connects your web scraping workflows directly to your AI client. Use the agent to list, launch, monitor, or retrieve data scraped from websites via natural conversation.
You can manage all Phantoms and workflows—from configuration checks to fetching final lead lists—without ever leaving your chat interface.
What your AI agents can do
Abort phantom
Stops any Phantom that is currently running.
Get phantom
Retrieves detailed information about a specific Phantom workflow.
Get phantom output
Fetches the final dataset and results from a completed scraping job.
Retrieves a list of every Phantom (scraping job) configured on your account.
Launches a specified Phantom, optionally passing custom arguments for the scrape.
Immediately terminates a Phantom that is currently executing.
Fetches the final, compiled dataset and output from a completed scrape.
Shows the exact configuration, arguments, and settings used for any specific Phantom.
Provides current account limits and usage statistics, letting you know how much execution time is left.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
PhantomBuster MCP Server: 10 Tools for Web Automation
Use these tools to manage every aspect of web scraping, from listing Phantoms to launching complex data extraction jobs.
019d75f2abort phantom
Stops any Phantom that is currently running.
019d75f2get phantom
Retrieves detailed information about a specific Phantom workflow.
019d75f2get phantom output
Fetches the final dataset and results from a completed scraping job.
019d75f2get phantom setup
Shows the specific configuration arguments used by a Phantom workflow.
019d75f2get phantombuster account
Gets your overall account usage limits and settings information.
019d75f2get workflow
Retrieves details for a specific, defined automation workflow.
019d75f2launch phantom
Starts a new Phantom scraping execution with specified arguments.
019d75f2list containers
Lists all currently active or recent Phantom containers running on your account.
019d75f2list phantoms
Returns a list of every saved and available Phantom scraping agent you own.
019d75f2list workflows
Shows all defined automation workflows available for use in your account.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with PhantomBuster, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
You're connecting your PhantomBuster account directly to your AI agent. This server puts full control over web scraping and lead generation right where you are chatting with your agent. You manage all of it without ever leaving this conversation window.
To see what agents you have, run list_phantoms, which gives you a list of every saved scraping job you own. If you want to know about specific automation workflows defined on the account side, use list_workflows to show all available options, and then get_workflow pulls up detailed info for any single workflow.
Need to start an operation? You launch a new scrape using launch_phantom, passing custom arguments if you need them. If the job is already running, run list_containers to see all active or recent Phantom containers on your account. To check the precise setup and configuration arguments used by any specific agent, run get_phantom_setup.
For deeper details about a particular scraping agent, call get_phantom. You can also pull up general usage limits and settings info for your whole account by running get_phantombuster_account.
If you gotta stop an active scrape mid-run because something's wrong or you changed your mind, run abort_phantom to immediately terminate the Phantom that’s currently executing. When a job is done and you need the goods, call get_phantom_output; this fetches the final, compiled dataset and results from the completed scrape.
If you're checking on an existing workflow definition, you can run get_workflow to retrieve its full details.
This setup lets your agent talk directly to PhantomBuster. You just tell your AI client what you want—like 'launch my LinkedIn export job with these parameters,' or 'show me all the data from that last scrape'—and the agent runs the required tool function, giving you the raw output right back in plain conversation.
How PhantomBuster MCP Works
- 1 Subscribe to the server and provide your PhantomBuster API Key.
- 2 Instruct your AI agent with a command (e.g., 'List all Phantoms and check the status of the LinkedIn export').
- 3 The agent calls the appropriate tool (
list_phantoms,get_phantom), and you receive the structured data output directly in the conversation.
The bottom line is, your AI client handles the API calls; you just talk to it.
Who Is PhantomBuster MCP For?
Growth marketers who spend hours manually checking which lead scraping jobs succeeded and which ones failed. Sales operations teams that need to ingest large batches of structured data into a CRM instantly. Data engineers debugging complex web collection pipelines.
Uses the agent to check active workflows (list_phantoms) and then retrieves the extracted lead data (get_phantom_output) for bulk CRM enrichment.
Needs to quickly launch a targeted scrape (via launch_phantom) on a new set of keywords or URLs, and immediately verify the output structure using get_phantom_setup.
Audits Phantom configurations (get_phantom_setup) to ensure the scraping logic hasn't broken after an update, then uses the agent to abort a failing container via abort_phantom.
What Changes When You Connect
- Need to see what's running? Call
list_phantomsto get a status overview of every scraping job without logging into the dashboard. It’s all in one chat command. - Launched a scrape but need to stop it fast? Use
abort_phantom. You can terminate any runaway Phantom directly through your AI client, saving usage time. - Got the raw data and need to check if the parameters were right? Run
get_phantom_setupfirst. Verify the arguments before you spend execution credits on a bad scrape. - Don't manually download files. Use
get_phantom_outputto pull the final, structured results directly into your agent's context for immediate analysis or processing. - Track your budget live. Call
get_phantombuster_accountto see your remaining execution time and account limits before starting a major automation batch.
Real-World Use Cases
The Cold Lead Follow-Up
A sales rep needs 50 names from a specific industry. Instead of building the scrape manually, they ask their agent: 'Launch my Industry X Phantom with these 10 keywords.' The agent calls launch_phantom, and once finished, they call get_phantom_output to get a clean CSV dataset ready for CRM import.
Debugging Broken Scrapes
A data engineer notices the LinkedIn scraper failed overnight. They first use list_phantoms to find the job ID, then call get_phantom_setup to check if a recent change broke the arguments. Finally, they can try rerunning it with adjusted inputs using launch_phantom.
Auditing Automation Rules
A marketing manager needs confirmation that their web scraping rules haven't changed since last week. They run get_workflow on the main lead-gen pipeline and confirm the exact parameters before approving a large, costly campaign launch.
Quick Status Check
The team needs to know which scrapes are running right now. They simply ask their agent to run list_containers. This instantly shows if everything is active or if something stalled out and needs the manual intervention of abort_phantom.
The Tradeoffs
Guessing current status
Trying to figure out if a Phantom is running, failed, or ready just by looking at old chat logs or documentation.
→
Always start with list_phantoms for an immediate overview. If you need the live status of containers, run list_containers. This gives you the definitive, real-time view.
Rerunning without checks
Launching a Phantom with the old parameters because it worked last month, only to find out the target website changed its structure.
→
Before launching, run get_phantom_setup. Compare those retrieved arguments against your current requirements. This prevents wasted execution time and bad data.
Manual data collection
Copying 50 leads from a website into a spreadsheet by hand after the scrape finishes.
→
Never copy-paste manually. Use get_phantom_output to pull the entire dataset directly into your agent's context, ready for immediate structured analysis or database upload.
When It Fits, When It Doesn't
Use this server if your main bottleneck is turning raw web content (like LinkedIn profiles or website listings) into usable data sets. You need a way to programmatically list, start, monitor, and retrieve the results of scraping jobs without touching a dashboard. Don't use it if you just need general automation—if you only need simple API calls like getting weather data, a standard REST client is enough. However, if your task involves complex web interaction (logging in, clicking through pages, extracting structured tables), then this server and its tools (launch_phantom, get_workflow) are the right fit because they manage those specific multi-step scraping processes.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by PhantomBuster. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Manually tracking leads across multiple websites is a nightmare.
Today, if you need to scrape 500 leads from three different sites, you run the job in the PhantomBuster dashboard. Then, you have to manually switch tabs every five minutes: checking which job finished, clicking 'View Output', and then copying data into a spreadsheet or CRM record. It's slow, it's error-prone, and it takes half your day.
With this MCP server, the process shifts. You tell your agent what to scrape using `launch_phantom`. The agent handles all the monitoring in the background. When the job is done, you simply ask for the results via `get_phantom_output`, getting a clean, structured data payload instantly. No clicks, no switching tabs—just talking.
PhantomBuster MCP Server: Get all scraping controls from chat.
Before this server, if you wanted to stop a runaway scrape or check what arguments were used, you had to navigate deep into the PhantomBuster UI and find the specific container ID. It was clunky and required multiple logins just for oversight.
Now, it's simple. You ask your agent: 'What is the status of my current scraping job?' The agent runs `list_containers` or `get_phantom`, giving you the precise answer in one step. Oversight becomes part of the conversation.
Common Questions About PhantomBuster MCP
How do I list all Phantoms using get_phantom? +
You use list_phantoms to see every Phantom agent you have saved. This tool provides a simple inventory so you know exactly what workflows are available for launching or checking.
I need to check my usage limits on the PhantomBuster MCP Server. +
Call get_phantombuster_account. This function retrieves your total account settings and tells you how much execution time is left, preventing you from accidentally running out of credits.
What tool should I use to start a scraping job? +
You need launch_phantom. Just give the agent the Phantom name and any required custom arguments. This sends the command to begin the scrape immediately.
How do I get the actual leads from a completed run? Use get_phantom_output. +
The get_phantom_output tool retrieves the final, extracted dataset for you. It gives you structured data ready to be used in your next step—no manual copy-pasting needed.
How do I diagnose why a specific Phantom failed using get_phantom? +
The tool retrieves detailed status information for any given Phantom. Look for the 'status' field; it tells you if the job succeeded or failed and often includes an error code explaining the failure.
Can I use get_phantom_setup to check the required arguments for a Phantom? +
Yes, this tool gives you the exact configuration needed before running. It lists all necessary inputs (arguments) and their expected data types so your agent knows how to structure the launch command.
How do I stop an active Phantom using abort_phantom? +
You simply call abort_phantom with the target's ID. This sends a kill signal, stopping any running process immediately and preventing further usage or billing accrual.
Where do I check my list of available automation workflows using list_workflows? +
This tool shows all defined multi-step processes (workflows), which are different from single Phantoms. It lets you see the high-level logic and sequence steps built into your overall automation suite.
How do I find my PhantomBuster API Key? +
In your PhantomBuster account, click on your name in the top right, go to Org settings or Workspace settings, and look for the API Key section.
Can I provide arguments when launching a Phantom? +
Yes! The launch_phantom action accepts an optional argument field where you can provide a JSON string of parameters to override the default Phantom configuration.
How do I see the data extracted by a Phantom? +
Use the get_phantom_output tool with the specific agent ID. It will return the results of the last successful execution of that Phantom.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
Synthesia
Produce professional AI videos with lifelike avatars that present in 140+ languages without cameras, studios, or actors.
HelpCrunch
Engage customers with live chat, email automation, and a knowledge base that reduces support workload and boosts satisfaction.
ON24 Virtual Events
Manage webinars and virtual events via ON24 — track registrants, attendees, and analytics directly from your AI agent.
You might also like
Jamf Pro
Manage Apple devices, computers, and inventory via Jamf Pro API.
AdRoll
Retarget customers, launch display ad campaigns, and measure conversion performance across web and social channels.
Transifex
Localize your projects efficiently using AI Agents with the official Transifex integration.