MCP Gtm Tech Stack Signal Scraper

What mcp-gtm-tech-stack-signal-scraper MCP server does

The mcp-gtm-tech-stack-signal-scraper MCP server takes a bare company domain and examines publicly accessible website pages and scripts for evidence of go-to-market software. Its output is designed for enrichment workflows that need a consistent row rather than a narrative response.

The returned fields classify detected tools by role, including CRM, sequencer, and marketing automation. The row also includes a total GTM tool count, a composite tech_stack_signal, and boolean fields such as uses_hubspot, uses_salesforce, uses_apollo, uses_gong, uses_intercom, and uses_marketo. This format can be passed into Clay, a CRM, or another agent workflow.

How it works

The MCP tool is named detect_gtm_tech_stack. It sends the domain and detection options to the Mamba Labs GTM Tech Stack Signal Enrichment actor on Apify. The package itself is a thin client; the website inspection and detection logic run through Apify.

The required domain value should contain only the company domain, such as stripe.com, without a protocol or trailing slash. By default, the scan can crawl up to two additional pages, such as pricing or product pages, when crawl_additional_pages is omitted. Set that option to control whether those pages are included.

Use technologies when the question concerns specific products rather than the full detectable set. Supported values include HubSpot, Salesforce, Marketo, Pardot, Intercom, Drift, Apollo, Outreach, Gong, and ZoomInfo, represented by lowercase identifiers. The mcp-gtm-tech-stack-signal-scraper MCP server then reports the requested technologies instead of every detectable one.

Setup and configuration

Running the mcp-gtm-tech-stack-signal-scraper MCP server requires Node.js 18 or later, an Apify account, and an Apify API token. Configure the server as a local stdio MCP process using the package published as @mambalabsdev/mcp-gtm-tech-stack-signal-scraper.

For Claude Desktop, add the package as an MCP server and provide the token through the APIFY_TOKEN environment variable. After restarting the client, the detect_gtm_tech_stack tool should be available. Each call consumes Apify credits, so usage depends on the Apify account and actor pricing rather than only on the local package.

Tools and capabilities

• Detect CRM, sequencer, and marketing automation signals from public websites.
• Return a flat JSON result with a stable set of fields.
• Report boolean indicators for HubSpot, Salesforce, Apollo, Gong, Intercom, and Marketo.
• Provide detected classifications, a GTM tool count, and a composite technology-stack signal.
• Restrict results to selected technologies with the optional technologies input.
• Crawl up to two additional pages for potentially broader website coverage.

Limitations and notes

Detection is based on public-facing website pages and scripts, so the output represents observable signals rather than a verified internal software inventory. The server is read-only and does not modify the target website or connected systems. It depends on Apify for execution and requires an APIFY_TOKEN; calls consume Apify credits. The complete actor schema, pricing, and run history are maintained on the referenced Apify Store page rather than in the MCP package itself.