AI EXTENSIONS: MCP SERVER & Skills

Deploy and run load tests,
just by asking AI.

Gatling AI Extensions connect your AI assistant to Gatling Enterprise Edition through MCP Servers & Skills. Configure, deploy, and start performance tests without leaving your IDE.

Try Enterprise Edition for free

Read Documentation

AVAILABLE FOR

Two components that bring Gatling Enterprise Edition into your coding agent.

Gatling AI Extensions ship two complementary components. Each solves a different part of the problem. Together, they take you from a blank project to a running test on Gatling Enterprise Edition, entirely
inside your AI assistant.

MCP Server

Connects your AI assistant to Gatling Enterprise Edition. Ask it about your infrastructure in plain language: teams, packages, simulations, load generator locations. All read from your real account, in real time.

AI Skills

/gatling-configuration-as-code & /gatling-build-tools

Guided, multi-step workflows for Claude Code. Skills orient your AI agent through the exact steps needed to configure and deploy a test, using the MCP server to pull live data from your account at each stage.

MCP Server

See the MCP server in action

Explore how you can query your tests and packages, find the right teams and
locations, and configure a deployment without ever opening the UI.

YOUR NEW END-TO-END WORKFLOW

Write a test.
Ask your coding agent to deploy it.

Write your load test

Author your load test in Java, Kotlin, Scala, or JavaScript, in your IDE.

Ask your coding agent to configure it

/gatling-configuration-as-code
‍
Your coding agent fetches your live account data, generates the config.

Ask your coding agent to deploy it

/gatling-build-tools
‍
Your coding agent deploys the package to Gatling Enterprise and captures all IDs.

Your test runs

Monitor results in Gatling Enterprise Edition.

Every step happened inside your IDE.

MCP SERVER

Give you coding agent full visibility into Gatling Enterprise Edition

The MCP server is a local server that runs on your machine and exposes your Gatling Enterprise Edition account to any MCP-compatible AI client.

No UI required. No manual lookups. Your AI can answer questions about your infrastructure the moment you ask them.

What your coding agent can access:

All teams in your Gatling Enterprise Edition account

Deployed packages, managed and private

All simulations and test configurations

Load generator locations, public and private

Runs entirely locally via stdio. Your API token and infrastructure data never leave your machine.

AI SKILLS

Production-ready deployment workflows, built into your coding agent.

Skills are structured instruction sets for Claude Code, designed to guide your AI agent through specific, multi-step tasks with built-in checks, correct parameters, and real data from Gatling Enterprise Edition.

They eliminate the back-and-forth that makes AI agents unreliable for technical workflows.

Skill 1: /gatling-configuration-as-code

The hardest part of getting a test onto Gatling Enterprise Edition is the config file. Wrong IDs, missing location weights, silent typos: it's tedious to get right manually. This skill handles it end-to-end.

The agent will:

Read your existing config file if one already exists

Pull your live teams, locations, packages, and tests from Gatling Enterprise via the MCP server

Ask you the right questions: which simulation, which team, which locations, what stop criteria

Write a valid, production-ready HOCON config with real IDs pre-filled

The result: a package descriptor ready for CI, built from your actual account data

Skill 2: /gatling-build-tools

This skill takes over once your config is ready. It detects your build tool, runs the right deploy command, captures the generated package and test IDs, and optionally kicks off the test run.

The agent will:

Detect your build tool: Maven, Gradle, sbt, npm/JavaScript

Verify your API token is set and has the correct permissions

Run the deploy command with the exact right flags for your build system

Update your config files with the generated package and test IDs

The result: your simulation is live on Gatling Enterprise Edition. All the steps you'd normally do through the UI, done by your AI agent.

Ready to deploy
your first load test
using AI agent?

Get started with the MCP Server and Skills. Your Gatling Enterprise Edition account is all you need.

Access MCP & Skills Documentation

FAQ

Frequently Asked Questions about Gatling AI Extensions

Do I need a Gatling Enterprise Edition account?

Yes. Both components connect to Gatling Enterprise Edition via an API token. You need a valid account and a token with the Configure role.

Does my data leave my machine?

No. The MCP Server runs locally via stdio. All communication is between your machine and your Gatling Enterprise Edition account directly.

Which coding agents are supported?

The MCP Server works with any MCP-compatible client: Claude Code, Claude Desktop, Cursor, Windsurf, VS Code with Copilot. Skills are packaged as a Claude Code plugin today and can be adapted for
other agents.