PhantomBuster
PhantomBuster is a cloud-based data extraction and automation platform that enables users to automate actions on the web, such as scraping data, generating leads, and automating social media tasks.
Verdict
Common use cases
- Scrape competitor LinkedIn pages on demand
- Export Instagram follower lists for outreach
- Pull Twitter mentions into weekly reports
- Harvest event attendee contacts from public sites
- Trigger lead enrichment agents from chat
Integration
- Vendor
- PhantomBuster
- Category
- other
- Auth
- API_KEY
- Tools
- 19
- Composio slug
phantombuster
Tools
- Create Branch
Tool to create a new branch. use when you need to isolate updates by creating a separate branch. use after authenticating with your phantombuster api key and optionally specifying an organization.
- Delete Branchdestructive
Tool to delete a branch by id. use when you need to remove obsolete or incorrect script branches; ensure the branch exists before calling.
- Export Agent Usage CSV
Tool to export agent usage csv for current organization. use when you need a downloadable report of all agents' run statistics in csv format.
- Export Container Usage CSV
Tool to export container usage csv for current organization. use when you need a downloadable csv report of container execution history up to 6 months; url expires in 30 days.
- Get Agent Groups
Tool to get agent groups and order for the current organization. use when you need to review how agents are grouped and ordered.
- Get All Agents
Tool to fetch all agents associated with the current user or organization. use after validating your phantombuster api key to list available agents.
- Get All Branches
Tool to fetch all branches associated with the current organization. use when you need to enumerate every branch across your scripts.
- Get All Scripts
Tool to fetch all scripts for the current user. use after authenticating your phantombuster api key to list scripts without their code bodies.
- Get Branches Diff
Tool to get the length difference between the staging and release branch of all scripts. use when assessing branch changes before deployment.
- Get Containers Fetch All
Tool to get all containers associated with a specified agent. use when you need to retrieve every container for a given agent, for monitoring or reporting.
- Get Deleted Agents
Tool to get deleted agents for the current user or organization. use when you need to audit recently removed agents.
- Get IP Location
Tool to retrieve the country of a given or environment ip address. use when you need to geolocate an ip endpoint. example: get location ip(ip="8.8.8.8")
- Get Organization
Tool to fetch current organization details. use when you need to retrieve the organization associated with the provided api key.
- Get Organization Resources
Tool to get current organization's resources and usage. use when you need to monitor quotas.
- Get Org Running Containers
Tool to get the current organization's running containers. use after listing agents to identify active containers across the organization.
- Get Org Storage Lists Fetch All
Tool to fetch all storage lists for the authenticated organization. use when you need to enumerate every storage list available (beta).
- Request AI Completion
Tool to request a text completion from the ai module. use when you need to generate text based on a prompt.
- Solve hCaptcha
Tool to solve an hcaptcha challenge. use when you need a valid hcaptcha token for form submissions or automation flows.
- Unschedule All Agent Launches
Tool to unschedule all scheduled launches for agents. use when you need to disable every automated agent run organization-wide after verifying your api key.
Setup
Setup guide
- 11. In Switchy, open Settings → Integrations → Add MCP Server and select PhantomBuster. 2. Log into your PhantomBuster account at phantombuster.com, navigate to Settings → API, and copy your API key. 3. Paste the key into Switchy's configuration field and click Connect. 4. Switchy will verify the key by fetching your agent list; you'll see a green checkmark when authentication succeeds. 5. Open any Space, type '@PhantomBuster get all agents' and send — the MCP will return your configured scraping agents. 6. To launch an agent, use '@PhantomBuster' followed by the agent name or ID; results appear in the Space once the job completes. 7. If you manage multiple organizations in PhantomBuster, specify the org ID in advanced settings to scope API calls correctly.
What teammates see: by default, memories from PhantomBuster are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
List Active Agents
@PhantomBuster get all agents in my account and show their names, statuses, and last run timesOpen in a Space →
Export Usage Report
@PhantomBuster export container usage csv for the current organization and provide the download linkOpen in a Space →
Create Script Branch
@PhantomBuster create a new branch called 'test-linkedin-v2' for script ID 12345 so I can iterate safelyOpen in a Space →
Review Agent Groups
@PhantomBuster get agent groups and their order for the current organizationOpen in a Space →
List All Branches
@PhantomBuster fetch all branches in the organization and display their names and creation datesOpen in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@PhantomBuster list all my active agents and show me which ones have run in the last 7 days
Based on your PhantomBuster account, you have 12 agents configured. In the past 7 days, 8 agents executed successfully: 'LinkedIn Profile Scraper' (47 runs), 'Twitter Follower Export' (23 runs), 'Email Finder' (19 runs), 'Instagram Engagement Bot' (15 runs), 'Sales Navigator Lead Gen' (12 runs), 'Facebook Group Scraper' (8 runs), 'Google Maps Extractor' (5 runs), and 'Crunchbase Company Enrichment' (3 runs). The remaining 4 agents ('YouTube Channel Analyzer', 'Reddit Post Monitor', 'Zillow Listing Scraper', 'Product Hunt Tracker') have not executed since their last manual trigger 14+ days ago.
This example demonstrates the read-heavy 'Get All Agents' capability. PhantomBuster's API returns agent metadata including execution history, but the depth of historical data depends on your subscription tier. Free plans may show limited run statistics compared to paid accounts.
@PhantomBuster create a new branch called 'linkedin-test-v2' for my LinkedIn scraper script so I can test profile extraction changes safely
Branch 'linkedin-test-v2' created successfully (ID: br_8x4k2m9p). This branch is now isolated from your production LinkedIn scraper. You can modify the extraction logic, adjust rate limits, or test new selectors without affecting live runs. The branch inherits your current script configuration but maintains separate execution history. Remember to merge or delete this branch once testing is complete to avoid confusion in your workspace.
This showcases the 'Create Branch' tool for version control workflows. PhantomBuster branches let you experiment with automation logic before deploying changes. You'll need write permissions on the parent script, and branches count toward your organization's script quota.
@PhantomBuster export a CSV of all container usage from the past 90 days and summarize which automations consumed the most execution time
Container usage CSV generated and available for download (expires in 30 days). Analyzing the past 90 days: Your 'Sales Navigator Lead Gen' agent consumed 47.3 hours of execution time across 312 runs, followed by 'LinkedIn Profile Scraper' at 31.8 hours (498 runs) and 'Email Finder' at 22.1 hours (1,047 runs). The top three agents account for 68% of your total container usage. Peak usage occurred during the first week of each month, correlating with your monthly lead generation campaigns. Consider optimizing the Sales Navigator script's wait times to reduce per-run duration.
This pairs PhantomBuster's 'Export Container Usage CSV' tool with AI analysis to surface resource consumption patterns. The CSV includes execution timestamps, duration, and agent IDs but requires manual correlation with agent names. Download links expire after 30 days, so save the file if you need long-term records.
Use-case deep-dives
When PhantomBuster wins for tracking scraper spend
A 3-person growth team runs 12 PhantomBuster agents scraping LinkedIn and Twitter for leads. Every Monday they need to see which agents burned the most container hours and whether any scripts failed silently over the weekend. The Export Container Usage CSV tool pulls six months of execution history in one call, and Get All Agents lists every scraper's last-run status without opening the web UI. This MCP is the right call if you're already paying for PhantomBuster and need to surface usage data in Slack or a shared doc. If you're not a PhantomBuster customer, this integration does nothing—it's a pure API wrapper for an existing subscription. The buying call: add this MCP when your team checks PhantomBuster dashboards more than twice a week.
Branch management for multi-agent workflows
A 6-person ops team maintains 19 PhantomBuster scripts that scrape competitor pricing, job boards, and review sites. When a script breaks in production, they need to roll back to a known-good branch without logging into the PhantomBuster UI. The Create Branch and Delete Branch tools let an AI agent snapshot working code, test changes in isolation, and clean up failed experiments—all from a Switchy chat. This MCP is borderline if your team only runs one or two agents; the branch overhead isn't worth it. The threshold: if you have more than five scripts and you've rolled back a change in the last quarter, this integration saves 10 minutes per incident. The buying call: adopt this when script stability matters more than launch speed.
When this MCP wins for client billing transparency
A 4-person agency manages PhantomBuster accounts for eight clients, each running custom lead-gen agents. At month-end, the agency needs to bill clients based on actual container usage and prove which agents ran when. The Export Agent Usage CSV tool generates a per-agent breakdown that maps directly to invoice line items, and Get Agent Groups shows how agents are organized by client. This MCP is the right call if you're reselling PhantomBuster capacity and need audit trails without manual dashboard exports. If you're an internal team with one account, the web UI's built-in reports are faster. The buying call: add this MCP when you manage three or more PhantomBuster organizations and bill usage back to stakeholders.
Frequently asked
What does the PhantomBuster MCP do in Switchy?
It lets your team create, manage, and monitor PhantomBuster automation agents directly from Switchy's AI workspace. You can list all agents, export usage reports, manage script branches, and check agent groups without switching to the PhantomBuster dashboard. Useful when you're building workflows that combine PhantomBuster scraping with other tools in your stack.
Do I need a PhantomBuster API key to connect this MCP?
Yes. The MCP uses API key authentication, so you'll need to generate a key from your PhantomBuster account settings. Any team member with a valid API key can connect the MCP, but they'll only see agents and scripts their PhantomBuster account has access to. Admin privileges in PhantomBuster aren't required unless you're managing organization-wide settings.
Can the PhantomBuster MCP launch or stop automation agents?
No. The 19 available tools focus on reading agent data, exporting usage CSVs, and managing script branches. You can't trigger agent runs, pause executions, or modify agent configurations through this MCP. For those actions, you'll still need to use PhantomBuster's web interface or their separate launch API endpoints.
Why use this MCP instead of PhantomBuster's dashboard directly?
The MCP makes sense when you're coordinating PhantomBuster data with other tools in Switchy—like pulling agent usage stats into a report alongside CRM data, or checking which scripts exist before writing new automation logic. If you're just running one-off scrapes, the PhantomBuster dashboard is faster. The MCP shines in multi-tool workflows.
Who on my team should connect the PhantomBuster MCP?
Whoever manages your automation scripts or needs to audit agent usage. Since the MCP uses an API key tied to a PhantomBuster account, that person's access level determines what the whole team sees in Switchy. If multiple people need different views, you'll need separate Switchy workspaces with different API keys.