MCP Company Contact Details Extractor

What mcp-company-contact-details-extractor MCP server does

The mcp-company-contact-details-extractor MCP server accepts a company domain and returns contact information published by that company. Its result is one flat JSON row with 27 snake_case fields rather than nested objects. The row can include role-based addresses for general inquiries, support, sales, and privacy; a main phone number; a postal address; the contact page URL; and extraction status details.

Named individual email addresses are excluded. Email addresses are retained when their domain matches the company domain and their local part belongs to the supported role vocabulary. Free-mail providers are rejected by default, but an option allows them for cases such as small businesses that publish Gmail or similar addresses.

How it works

The package is a thin client for the Mamba Labs Company Contact Details Extractor actor hosted on Apify. The mcp-company-contact-details-extractor MCP server starts an actor run, polls until it reaches a final status, and returns the actor result without applying its own extraction logic.

The lookup begins with the supplied bare domain, such as stripe.com. The actor finds a contact page and can read a second support or legal page when configured with a crawl depth of 2. Phone extraction looks for a main number and can format it as E.164 when the country is resolvable. Address extraction checks JSON-LD before the page footer.

Successful lookups are cached for seven days unless skipCache is enabled. The response includes fields such as contact_page_method, fetch_status, coverage, pages_read, and counts and reasons for rejected email addresses.

Setup and configuration

Run the server with Node.js 18 or newer and an Apify account with an API token. A Claude Desktop configuration can launch it through npx using the package @mambalabsdev/mcp-company-contact-details-extractor. Set the required APIFY_TOKEN environment variable, then restart the client.

The required input is company_domain. Optional inputs include company_name, emailTypes, includePhones, includeAddress, allowFreeMailboxes, crawlDepth, and skipCache. emailTypes accepts all, general, support, sales, privacy, general_support, or sales_only.

Tools and capabilities

The server exposes the extract_company_contact_details tool. It can:

• Find a company contact page from a domain.
• Return role-based general, support, sales, and privacy emails.
• Extract a main phone number and postal address when published.
• Report the contact page URL and how it was located.
• Return rejection counts and reasons for excluded addresses.
• Produce a flat row suitable for downstream Clay workflows.

Limitations and notes

The mcp-company-contact-details-extractor MCP server depends on Apify for all extraction and requires a valid API token. It does not return named-person email addresses. A missing or invalid required domain produces an error before the actor run starts.

Runs can continue for a long time. The actor is allowed up to 1,800 seconds; if it remains in progress two minutes beyond that limit, the call stops waiting and returns the run ID and an Apify Console link. Failed runs return an error with their run ID and status. Narrow email-type filters leave other email columns null, which differs from proving that an address was not found.