# BlazeMeter MCP for AI Agents AI Agent Connect > BlazeMeter MCP lets you manage load testing workflows directly from your AI client. You can list workspaces and projects, trigger performance tests, monitor live run metrics like p90 and p99 KPIs, and shut down active cloud connections. It brings enterprise-grade load testing into your chat interface for faster QA and SRE cycles. ## Overview - **Category:** ship-it - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_r6TJf0A2rRQHYgLlCVZI7oB2MYzqNtdD2bTXNTMV/ai-agent-connect - **Tags:** performance-testing, load-testing, stress-testing, continuous-testing, qa-automation, infrastructure-monitoring ## Description This BlazeMeter MCP lets you manage load testing workflows directly from your AI client. Managing performance tests usually means jumping between several browser tabs, checking load distributions, and manually triggering JMeter definitions. This Connector changes that by letting your AI agent handle the heavy lifting. You can ask your agent to pull a report on a specific test run or find the right project for a new stress test. It brings your load testing data into a single conversation. Because Vinkius hosts this Connector, you can connect it to your preferred AI client and start interacting with your BlazeMeter environment immediately. You can check your infrastructure health, start new load tests, and even kill runaway processes using plain English. It moves the complexity of JMeter and cloud scaling into a simple back-and-forth dialogue. Instead of logging into a dashboard to check if your p99 metrics are stable, you just ask your agent to show you the latest numbers. This keeps you in your flow state while ensuring your production environment stays stable. You get direct control over cloud-based performance hosts without the friction of manual configuration. This connection bridges the gap between your testing infrastructure and your daily communication habits. You can manage workspaces, extract project structures, and monitor live runs with the same ease as sending a message to a teammate. It turns a complex infrastructure operation into a simple query and response cycle. ## Tools ### stop_master Shutdown active gateway arrays. This is your emergency brake for runaway load tests. ### get_report View internal arrays for specific plan math reports. This provides the deep data you need for analysis. ### get_user Identify specific user records in the identity parsing system. Use this to manage account-level data. ### list_workspaces See all your bounded workspaces in the BlazeMeter platform. This helps you navigate large enterprise environments. ### list_projects Extract a list of projects tied to a specific workspace. Use this to find the right test group quickly. ### list_tests Get a JSON payload of all tests in a project. This is useful for seeing what configurations are available. ### get_test Pull the specific configuration for a single test. Use this to check vault limits and test details. ### start_test Execute a load generation test and see the resulting metrics. This starts the actual cloud scaling process. ### list_masters See the active master records for your runs. This helps you track which orchestrators are running. ### get_master Check the status of a specific gateway run. Use this to see if a master is healthy. ## Prompt Examples **Prompt:** ``` List the performance testing projects inside Workspace ID 123456. ``` **Response:** ``` I queried your BlazeMeter workspace. Here are the active projects discovered: - **Core Backend API Stress** (ID: 9942) - **Frontend Cache Bypass Load** (ID: 9945) ``` **Prompt:** ``` Trigger a new execution for load Test ID 987654. ``` **Response:** ``` Understood. I executed a command to start the test. The BlazeMeter orchestration engines are now spinning up the performance nodes (Master ID: **m-11223**). You can ask me for status updates anytime. ``` **Prompt:** ``` Stop the actively running Master test ID m-11223 immediately. ``` **Response:** ``` Emergency stop validated. I have sent the shutdown instruction to the load orchestrator (**m-11223**). All active connections are being severed now. ``` ## Capabilities ### Trigger cloud performance tests Start new load generation runs and spin up cloud-based performance hosts using natural language. ### Monitor live p90 and p99 metrics Get real-time updates on your throughput and latency KPIs without leaving your chat. ### Kill runaway load tests Shut down active cloud connections immediately to protect your source architecture from overload. ### Fetch detailed performance reports Pull deep internal arrays and plan math reports for specific test runs. ### Manage BlazeMeter projects List and extract project metadata and JMeter definitions across your workspaces. ### Track active master records See exactly which orchestrators are currently running and check their gateway status. ## Use Cases ### Emergency shutdown of a runaway test A production server starts spiking. An SRE asks the agent to use stop_master to kill the active test immediately. ### Quickly checking p99 metrics A QA engineer finishes a load test and asks the agent to use get_report to see if the latency stayed under 200ms. ### Finding project configurations A DevOps engineer needs to see what tests are available for a new microservice and asks the agent to use list_tests. ### Monitoring live run health An SRE wants to know if a master run is still healthy and asks the agent to use get_master to check the gateway status. ## Benefits - Stop context-switching by using list_projects and list_tests to find your test configurations without opening a browser. - Get immediate visibility into performance with get_report to see p90 and p99 metrics as soon as a test finishes. - Protect your infrastructure by using stop_master to kill runaway cloud connections with a simple voice or text command. - Accelerate your QA cycles by using start_test to trigger cloud-based performance hosts on demand. - Simplify workspace management by using list_workspaces to navigate complex enterprise environments using natural language. - Monitor live run health using get_master to verify gateway status during critical testing windows. ## How It Works The bottom line is you get to control your entire performance testing suite through a natural conversation. 1. Subscribe to the BlazeMeter MCP on Vinkius. 2. Enter your BlazeMeter Key ID and Secret into your AI client's configuration. 3. Ask your agent to list projects, start a test, or check live metrics. ## Frequently Asked Questions **Can I use BlazeMeter MCP to stop a test that's crashing my site?** Yes. You can use the stop_master capability to immediately shut down active cloud connections and kill runaway load tests using a simple chat command. **How do I see the p99 results of my last load test?** Just ask your agent to pull the report for that specific test. It will fetch the internal math and KPIs like p90 and p99 directly into the chat. **Can I trigger a new test with a chat command?** Yes. You can tell your agent to start a specific test ID, and it will trigger the cloud-based performance hosts for you. **Does BlazeMeter MCP support JMeter?** Yes. This Connector can interact with your JMeter definitions, allowing you to list and execute those tests through your AI client. **How do I find my project IDs in BlazeMeter?** You can ask your agent to list the projects in your workspace. It will provide a list of all available projects and their corresponding IDs. **Can I monitor multiple test runs at once?** Yes. You can ask your agent to list all active master records to see the status of various running tests across your environment. **Can my agent actually start a cloud performance test orchestrating servers?** Yes. By utilizing the 'start_test' tool paired with an active Test ID, your AI agent securely pulls levers directly on BlazeMeter's cloud infrastructure to provision load generation hosts without any required GUI interaction. **How do I check the live status of an active test run (Master)?** You can instruct your agent to use the 'get_master' or 'get_report' tool. It strictly evaluates live Gateway stats, throughput rates, and p90 indicators dynamically reporting back. **Is there a way for the AI to forcefully kill a rogue load test?** Absolutely. Your agent can execute the 'stop_master' mutation—which immediately severs the test execution pipeline and protects vulnerable origin servers from unplanned stress.