Skip to main content
The LoneScale MCP server lets AI assistants (Claude, Cursor, ChatGPT, custom agents) call LoneScale’s enrichment and sourcing capabilities as tools — directly from a chat, no glue code required. It implements the Model Context Protocol over Streamable HTTP and authenticates users via OAuth 2.0 with PKCE. Behind the scenes, every tool call uses the same LoneScale Public API documented elsewhere on this site.

​
Endpoint

https://mcp-enrich.lonescale.com/mcp
Transport: Streamable HTTP (MCP spec ≥ 2025-06-18). Auth: OAuth 2.0 (user-consent flow — see Connect to Claude).

​
Available tools

All four tools below are available once the user has authorized with their LoneScale API key.
ToolModeDescription
public-api-companies-searchsyncLook up a company by name, domain, LinkedIn slug, or LinkedIn ID. Returns one fixed-shape company profile.
public-api-enrichasyncTrigger waterfall enrichment (email and/or phone) for a list of contacts. Returns a lonescale_job_id.
public-api-contact-sourcingasyncSource contacts at a company by persona, seniority, and location. Returns a lonescale_job_id.
get-job-resultsyncPoll for results of an async job using its lonescale_job_id. Returns pending until status is ready.
Sync tools return results inline. Async tools return a job id immediately and the assistant must call get-job-result every few seconds until the status is ready.

​
What you’ll need

  • A LoneScale account with an API key (Dashboard → Integrations → Public API).
  • A client that supports MCP custom connectors — for example Claude, Cursor, or any agent built with the Anthropic SDK.

Connect to Claude

Step-by-step setup, including the OAuth consent screen.

Examples

Three concrete prompts you can copy-paste into Claude.