You're probably already asking Claude or ChatGPT to draft an email, summarize a call, or research an account. What most people haven't done yet is let that same chat window actually touch their pipeline: pull a prospect list, enrich it, and hand it back ready to sequence, without opening Enginy at all.

That's what Enginy's MCP does.

Model Context Protocol is an open standard that lets an AI assistant like Claude or ChatGPT talk directly to a tool's real data and actions, with no copy-pasting between tabs and no exporting CSVs. You ask in plain English; the tool does the work and hands the result back into the same chat.

Once it's connected, Claude or ChatGPT can reach into your actual Enginy workspace: search and build contact or company lists, enrich and verify records, draft AI-personalized messages, create or update campaigns, and check performance, all as one conversation instead of five browser tabs.

If you're the kind of person who already lives in Claude or ChatGPT for research and drafting, this isn't a new tool to learn; it's your existing habit getting a lot more useful. Here's what that looks like in practice.

Enginy runs this through AI Finder and shows you a preview first: nothing gets imported into your workspace until you approve it. If the first pass is close but not quite right, just tell it what to adjust ("drop anyone below Director level") before you commit.

Tip: the more specific you are on seniority, geography, and company size, the less cleanup you do later. Same rule as any list-building tool.

One honest number worth knowing: verifying an email costs 1 credit; enriching a phone number costs 40. Both are one line in a chat, but they're not the same investment. Save phone enrichment for the accounts you're actually prioritizing, not the whole list by default.

This pulls from the real fields on each contact and company record, not a generic mail-merge; it's the same AI messages and variables you'd otherwise build by hand inside Enginy.

One caveat worth knowing up front: this only works cleanly on a draft campaign. If the campaign's already live, Claude/ChatGPT will need to clone it first before editing. It's the same safeguard that stops you from accidentally editing something that's already sending.

No dashboard required. You get the same answer you'd have gone digging for in Analytics.

Once the five prompts above feel routine, a few combinations worth trying: chaining enrichment → AI variables → campaign entry into a single Enginy workflow so new contacts get treated consistently without you re-typing the same three prompts every time; and connecting a webhook so your team gets a Slack ping the moment a high-intent reply comes in, instead of someone having to go check.

Requirements: Active Enginy account. Any Claude plan, including Free (Team and Enterprise plans need an org Owner to add the connector first). Any ChatGPT plan, including Free (Enterprise needs an admin to grant access first). No API keys, no technical setup beyond that.

Every MCP action draws from the same Enginy credit balance as the in-app UI. There's no separate charge for going through Claude or ChatGPT. Two habits keep that from surprising you.

Add "what could this cost?" before a bulk action (enrichment, AI variables, imports). Claude or ChatGPT will estimate the credit cost based on how many records and which action, so you get a cost-check before it processes, not after.

An estimate isn't the final number. After a run finishes, check the official Usage page to confirm actual credit consumption, especially for pricier actions like phone enrichment (40 credits) where a wrong list size adds up fast.

Not every model handles MCP equally well. As a rule of thumb:

⚠️

Not recommended: Claude Haiku, GPT Nano

These models are optimized for speed and low cost, not complex reasoning. They tend to misuse tools, miss steps, or produce incorrect outputs when working with MCP.

AI tools are genuinely good at thinking: research, drafting, pattern-spotting. But thinking doesn't ship pipeline on its own. Enginy is what takes the work AI starts and keeps it running: every prospect tracked through reply, follow-up, and scoring, synced into your CRM, your email stack, and your compliance setup, continuously, not just in the minute you happened to be prompting it.

With the skills’s Library, you don't need to write these prompts from scratch each time. Ask Claude or ChatGPT for the saved skill by name and it'll run the same workflow with less typing.

No new bill. Actions run against your normal Enginy credit balance, same as if you'd clicked the button inside the app.

No. Claude/ChatGPT is the interface; your CRM stays the source of truth and Enginy is what runs the work in between.

No. Plain, specific English works fine. The examples above are the actual level of detail to aim for.

Functionally yes. Connect using the ChatGPT link above and the same prompts apply.

Related reading: see the Help Center article on Enginy MCP for setup troubleshooting and platform-specific details.