Gatling MCP Server

The Gatling MCP Server exposes your Gatling Enterprise resources to AI assistants via the Model Context Protocol, enabling natural language interaction with your load testing infrastructure.

The Gatling MCP Server bridges AI applications with Gatling Enterprise, enabling AI agents to interact with your load testing infrastructure through natural language. Query teams, packages, simulations, and load generator locations directly from your AI assistant.

Installation#

Prerequisites#

You need a Gatling Enterprise account and an API token. Tokens can be generated from the Gatling Enterprise UI under API Tokens.

In all our examples, the API token is provided via the GATLING_ENTERPRISE_API_TOKEN environment variable.

Depending on your chosen installation method, you also need:

• Node.js v24 or later (includes npx)
• Docker installed and running

Usage with Claude#

The simplest way to use the Gatling MCP Server is to use the Gatling AI extensions plugin for Claude. However, you can also install it manually by adding the MCP server to the appropriate configuration files.

We recommended using Claude CLI as it is usually better integrated into IDEs (IntelliJ IDEA, VS Code…).

Claude CLI#

These configurations assume that you have the GATLING_ENTERPRISE_API_TOKEN environment variable set in your shell.

Using NPX:

claude mcp add gatling \
  --env 'GATLING_ENTERPRISE_API_TOKEN=${GATLING_ENTERPRISE_API_TOKEN}' \
  -- npx -y @gatling.io/gatling-mcp-server

Using Docker:

claude mcp add gatling \
  --env 'GATLING_ENTERPRISE_API_TOKEN=${GATLING_ENTERPRISE_API_TOKEN}' \
  -- docker run --rm --interactive --env GATLING_ENTERPRISE_API_TOKEN gatlingcorp/gatling-mcp-server

Claude Desktop#

Add the following to your claude_desktop_config.json:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json
• Linux: ~/.config/Claude/claude_desktop_config.json

Using NPX:

{
  "mcpServers": {
    "gatling": {
      "command": "npx",
      "args": ["-y", "@gatling.io/gatling-mcp-server"],
      "env": {
        "GATLING_ENTERPRISE_API_TOKEN": "<your-api-token>"
      }
    }
  }
}

Using Docker:

{
  "mcpServers": {
    "gatling": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "--interactive",
        "--env", "GATLING_ENTERPRISE_API_TOKEN",
        "gatlingcorp/gatling-mcp-server"
      ],
      "env": {
        "GATLING_ENTERPRISE_API_TOKEN": "<your-api-token>"
      }
    }
  }
}

Usage with VS Code#

Add the following to your VS Code settings.json or .vscode/mcp.json:

Using NPX:

{
  "servers": {
    "gatling": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@gatling.io/gatling-mcp-server"],
      "env": {
        "GATLING_ENTERPRISE_API_TOKEN": "<your-api-token>"
      }
    }
  }
}

Using Docker:

{
  "servers": {
    "gatling": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "--interactive",
        "--env", "GATLING_ENTERPRISE_API_TOKEN",
        "gatlingcorp/gatling-mcp-server"
      ],
      "env": {
        "GATLING_ENTERPRISE_API_TOKEN": "<your-api-token>"
      }
    }
  }
}

Tools#

The MCP server exposes the following tools, grouped by resource:

Teams#

• teams.read_all: List all teams visible to the API token.

Packages#

• packages.read_all: List all managed and private packages visible to the API token.

Source repositories#

• source_repositories.read_all / source_repositories.read_one: List or get the details of registered git repositories a test can build from.
• source_repositories.create_one: Register a new source repository.
• source_repositories.delete_one: Delete a source repository.

Locations#

• locations.read_all: List all managed and private locations where tests can be run.

Tests#

• tests.read_all / tests.read_one: List or get the details of tests.
• tests.create_one: Create a new test from a package or source repository.
• tests.patch_one: Update a test’s configuration.
• tests.delete_one: Delete a test.
• tests.start_one: Start a run for a test.

Runs#

• runs.read_all / runs.read_one: List or get the details of runs.
• runs.read_status: Get a run’s current status.
• runs.read_run_logs: Get a run’s execution logs.
• runs.read_report_requests / runs.read_report_groups: Get a run’s per-request or per-group performance statistics.
• runs.create_public_link: Create a public, unauthenticated link to a run’s dashboard.
• runs.patch_one: Update a run’s details.
• runs.stop_one: Stop a running run.

Each tool’s own description states its required role (Read, Configure, or Start) and any other constraints. See the gatling-mcp skill below for the concepts and workflows that tie them together.

Skill#

When installed via the Gatling AI extensions plugin, the MCP server is paired with the gatling-mcp skill, which documents these tools’ terminology (tests, runs, sources, packages, locations, teams), required roles, and recommended workflows so the AI assistant uses them correctly.

Example use cases#

Once connected, you can drive Gatling Enterprise entirely through natural language. A few things to try:

Clean up tests#

Rename all my tests that still use the old naming convention.

Lists your existing tests, then updates the ones that match.

Register a new test#

Create a test for this repo, build it with Maven, and run it from the Paris location.

Registers the source repository if needed, then creates a build-from-sources test. If the project isn’t git-buildable, pair this with the gatling-build-tools skill to package and upload the artifact first, then create a test pointing at the resulting package instead.

Diagnose a failed run#

My last run failed, what happened?

Pulls the run’s logs and cross-references its performance report to tell you whether it was a build error, an injection crash, or a stop criteria breach.

Analyze results#

What's the p95 response time for the login request in the last run?

Reads the run’s per-request report and pulls out the matching request.

Edit this page on GitHub