Trykitt MCP server

@pipeworx/trykitt

Find a verified work email by full name + company domain, via TryKitt.ai. Part of the enrich-waterfall finder rotation (fleet #2860).

Part of Pipeworx — an MCP gateway connecting AI agents to 1745+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Tools

confidence (0-1), and sources, requested in realtime mode so the HTTP response carries the finished result rather than a job id to poll.

• trykittfindemail(fullname, domain, linkedinurl?, _apiKey) — email,

Auth

BYO only. Pass your TryKitt API key as apiKey (header x-api-key — not** Authorization, which several third-party write-ups wrongly assume). Get one at https://trykitt.ai. No platform key (D4).

Data sources

• https://api.trykitt.ai/job/find_email — email finder (realtime mode).

Verification notes (fleet #2860, 2026-10-08, no vendor key held)

distinction: no auth header -> HTTP 401 {"message":"Auth credential not set in request","responsecode":401}; x-api-key: invalidkey123 -> HTTP 401 {"message":"Invalid API Key","response_code":401}. Two different messages for two different failures confirms the header is actually being read, not just any 401 default.

• **Header name confirmed by edge probe, including the "missing vs wrong"

third-party MCP wrapper READMEs (not TryKitt's own prose docs, which sit behind the same auth wall as the API) — we hold no key, so no successful call was made. Documented shape: { jobid, status, result: { email, confidence, sources } }. This pack treats any status other than "completed" as a loud error rather than a silent no-match, since TryKitt's job model can also run async — if a live key ever gets back a "pending" status despite realtime: true, that needs a poll step added, and the current code will say so explicitly instead of reporting found: false.

• **Response shape is from TryKitt's public Postman collection and

a synchronous result on every plan tier; the full set of status values.

• Not verified (no key held): whether realtime: true reliably forces

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "trykitt": {
      "url": "https://gateway.pipeworx.io/trykitt/mcp"
    }
  }
}

What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/trykitt/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — askpipeworx, discovertools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1745+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

Shortened. The full README is on GitHub.