MCP and agents

How to use the Airscale MCP (+Claude demo)

Want to prospect without ever leaving your AI chat? This tutorial covers how to use the Airscale MCP: the feature that connects your AI app (like Claude or ChatGPT) directly to Airscale's search, enrichment, and export tools.


What is the Airscale MCP?

The Airscale MCP connects your AI app (Claude, ChatGPT, etc.) directly to Airscale's search, enrichment, and export tools. This means you can prospect entirely through conversation, without switching between platforms.

Once the MCP is connected, you can:

  • Check your Airscale credits.

  • Search for people and companies.

  • Discover accepted Find Companies filter values.

  • Enrich contacts with professional emails, phone numbers, personal emails, and profile data.

  • Run AI web research with Airsearch.

  • Export people or companies to CSV or JSONL.The live MCP server currently exposes 19 tools, covering the main Airscale features.

Setup instructions are available on the dedicated AirSchool page.

How credits work

Credit consumption through the MCP mirrors the public API endpoints:

  • Most search and enrichment features consume the same amount of credits as the public API

  • Find People and Find Companies cost 0.1 credit per result

This ensures full transparency and consistency between the MCP and the rest of the Airscale platform.

Run a search in natural language

Once the MCP is set up in your LLM, you can open your app and simply describe what you want to do.

In this example, the prompt is:

"Find cybersecurity companies in France with 51 to 200 employees and upload them to a CSV file."

Adding the CSV instruction is useful because it triggers a split view, showing the results in a table next to Claude's response. This makes it easier to visualize the data and see live updates when enrichments are added later.

Claude then automatically uses the Airscale connector to run the search. Depending on your MCP permissions setup, Claude may:

  • Ask for permission before running each action, or

  • Run everything automatically (if all permissions have been pre-approved)

Review your results

Once the search is complete, Airscale returns a fully structured table. In this example, 99 companies were found using the Find Companies filters.

The table automatically includes key columns such as:

  • Company name

  • Domain

  • Website

  • Country

If the split view feels narrow, you can click Show in folder to open the full table as a CSV document.

Enrich contacts through conversation

From here, you can continue the workflow entirely through natural language. For example:

"Pull the CEO, CFO, and Head of Growth for each company and update the CSV."

Claude then triggers the appropriate Airscale enrichment tools and updates the table with:

  • Decision-makers' names

  • Job titles

  • LinkedIn handles

You can then keep going by asking for professional emails, phone numbers, or any other data point needed for outreach.

Export your final list

At any point, you can open the full table via Show in folder to access your CSV document, ready to be imported into your CRM, cold email platform, or any other tool in your stack.

Get started

To set up the Airscale MCP and start prospecting directly from Claude or ChatGPT, check out the dedicated AirSchool setup guide. If you need any help, the Airscale team is always available to support you.

On this page

© 2026

Airscale · All rights reserved