# Gatling MCP > Gatling MCP manages complex performance testing cycles right from your AI client. You can list simulation scenarios, kick off new load runs, track Virtual User counts as they spike, and pull detailed metrics like request error rates—all through natural conversation. It lets you manage everything from team quotas to resource pools without touching a dashboard. ## Overview - **Category:** developer-tools - **Price:** Free - **Tags:** load-testing, performance-testing, simulation, stress-testing, automation, ci-cd-integration ## Description Performance testing used to mean opening the dedicated platform, navigating deep into menus just to start a test run or check if it failed halfway through. Now, your agent handles that whole workflow for you. You tell it what load scenario needs running—maybe 'Search-API-Performance'—and it triggers the job on Gatling Enterprise infrastructure. Then, instead of refreshing pages and hunting down run IDs, you simply ask it to track progress or pull detailed stats like total requests and error counts. It’s about taking full control of your high-scale load simulations using plain language. By connecting your account through Vinkius, you gain access to a complete set of tools that covers everything from auditing resource pools to stopping an overrunning test immediately. This capability lets QA engineers and DevOps teams manage the entire performance lifecycle in one chat window. ## Tools ### list_simulations Lists all active load scenarios on Gatling Enterprise, providing their IDs, names, and associated teams. ### get_simulation Retrieves complete details for a specific Gatling simulation scenario. ### start_simulation Initiates a new performance test run on the Gatling Enterprise infrastructure and returns the unique run ID. ### abort_simulation Immediately halts any running Gatling simulation to manage resources or prevent overruns. ### list_runs Retrieves a list of past and active runs for a given simulation ID. ### get_run Fetches the complete details, status, and metrics for a specific test run. ### list_teams Lists every team registered within your Gatling Enterprise account. ### list_packages Lists all uploaded test packages or artifacts, helping you verify versions and upload dates. ### list_tokens Displays existing API tokens configured within Gatling Enterprise. ### list_pools Retrieves a list of available load generator pools, showing regions and instance counts for scaling checks. ## Prompt Examples **Prompt:** ``` List all simulations on Gatling Enterprise ``` **Response:** ``` Retrieving simulations... I found 3 active scenarios: 'User-Login-Stress-Test', 'Search-API-Performance', and 'Checkout-Flow-Baseline'. Which one would you like to start or view history for? ``` **Prompt:** ``` Start simulation 'abc-123' ``` **Response:** ``` Simulation run started! I've triggered 'abc-123' on the Enterprise infrastructure. Your run ID is 'run_xyz789'. I'll monitor the progress and notify you when the virtual user peak is reached. ``` **Prompt:** ``` Show me the stats for run 'run_xyz789' ``` **Response:** ``` Retrieving stats for run_xyz789... The test achieved a peak of 500 Virtual Users. Total requests: 12,500. Error rate: 0.05%. Average response time: 120ms. Would you like a breakdown by request type? ``` ## Capabilities ### Kick off load tests Start new Gatling simulations on the Enterprise platform, getting a unique run ID back instantly. ### Monitor and audit runs Track the progress of running or finished tests, including peak Virtual User counts and overall execution status. ### Stop runaway simulations Immediately abort a live load test run to save system resources when something goes wrong. ### Review detailed metrics Retrieve full test statistics, including request counts, error rates, and start/end times for deep analysis. ### Manage team capacity List registered teams and check member counts or credit quotas to ensure you don't hit usage limits. ## Use Cases ### Need to check if the new checkout flow is stable? The QA engineer asks their agent: 'List simulations' to confirm the correct scenario ID, then runs `start_simulation` with that ID. Finally, they ask for stats on a specific run ID using `get_run` to verify the error rate is below 0.1%. ### The deployment team needs to know if we have enough capacity. A DevOps lead uses the agent to call `list_pools`. This confirms that the required regions are available and lists the current instance counts, allowing them to validate scaling readiness before a major release. ### We suspect a test run is going rogue and consuming too many credits. An Ops Manager immediately tells their agent: 'Abort that simulation.' This uses `abort_simulation` to cut power to the generators, preventing accidental credit overruns. ### The team needs to plan for next quarter's peak traffic. A Performance Architect uses the MCP to run a baseline test, then uses `list_teams` and `get_run` repeatedly to gather metrics and ensure the current allocated quotas can handle projected growth. ## Benefits - Manage the entire test lifecycle without leaving your chat window. Need to start a run? Just ask your agent, and it triggers the job on Gatling Enterprise. - Stop wasting time checking dashboards. You can track progress or retrieve full details for any test run using the `get_run` tool, giving you immediate status updates. - Audit capacity easily. Check resource limits by listing load generator pools (`list_pools`) and verifying team quotas (`list_teams`)—all done with a simple prompt. - Control your environment instantly. If a test runs too long or hits an unexpected spike, use `abort_simulation` to shut it down immediately. - Deep dive into results. Beyond just pass/fail, you can get detailed stats for any run, checking request counts and error rates crucial for debugging. ## How It Works The bottom line is you manage complex load testing workflows through simple chat commands rather than navigating multiple web dashboards. 1. Subscribe to this MCP on Vinkius and enter your Gatling Enterprise API Token. 2. Connect your AI client (like Cursor or Claude) to the MCP endpoint using that token. 3. Use natural language prompts with your agent to list simulations, start a test run, or request performance statistics. ## Frequently Asked Questions **How do I start a test run with the Gatling MCP?** You use the `start_simulation` tool. You just need to tell your agent which simulation scenario you want to run, and it handles triggering the process. **Can I check my team's credit usage with the Gatling MCP?** Yes, you can use `list_teams` to view registered teams. This helps you monitor member counts and verify quotas before running large load tests. **What if a test run is going too far? How do I stop it?** If you need to halt an active simulation, use the `abort_simulation` tool. This stops the generators immediately, saving resources and preventing overruns. **Does Gatling MCP only show success or failure? What about metrics?** No, it provides deep metrics. After using `get_run`, you get full stats including total requests, error counts, and average response time for detailed debugging. **How do I see what test packages are available to use?** You can list your artifacts by calling `list_packages`. This shows the names, versions, and upload timestamps of all uploaded materials.