# LoadNinja MCP for AI Agents AI Agent Connect

> LoadNinja (Real-Browser Load Testing) lets you manage your performance testing lifecycle through your AI client. Trigger real-browser load tests, monitor live metrics, and audit virtual user (VU) limits without switching tabs. It connects your LoadNinja account directly to your agent for instant scenario orchestration and infrastructure auditing.

## Overview
- **Category:** developer-tools
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_flwAteXzuZIz58RcBrW0pK5weKMRXmdJ8NviFra4/ai-agent-connect
- **Tags:** load-testing, performance-engineering, browser-automation, scalability, stress-testing

## Description

Imagine you're in the middle of a high-stakes release and the staging environment starts feeling sluggish. Instead of hunting through a complex dashboard to find out where the bottleneck is, you just tell your AI agent to pull the latest stats for your current run. You get a clear breakdown of throughput and P95 transaction times immediately. If a test goes haywire and starts eating up resources, you don't have to scramble to find the stop button; you just tell the agent to kill it. This Connector bridges the gap between your performance engineering goals and your daily chat interface. You can check which data centers are available for your next global rollout or see exactly how many virtual users your subscription supports, all while staying in your workspace. It turns a multi-tab manual process into a conversation. Since Vinkius hosts this in their catalog, you can get it running in minutes and start managing your load testing infrastructure without the usual friction of manual report exports or constant context switching.

## Tools

### get_scenario
Pull the full configuration and target URL for a specific test scenario.

### run_scenario
Start a specific load test with a set number of virtual users and a duration in minutes.

### list_test_runs
See a list of all your past and currently active test executions.

### get_test_run
Get the full summary and details for a specific completed test run.

### get_test_run_stats
Pull the raw performance data and statistics from a specific test run.

### stop_test_run
Instantly terminate an active load test to free up resources or stop an error.

### list_browsers
See which native browser types are available for your load tests.

### list_scenarios
Get a complete list of all the load test scenarios currently saved in your account.

### get_account
Check your LoadNinja subscription details and see your current virtual user limits.

### list_locations
View all the physical data center locations where you can inject load.

## Prompt Examples

**Prompt:** 
```
List all my load test scenarios.
```

**Response:** 
```
I've found 4 scenarios in your account:

| Scenario Name | ID | Target URL |
| :--- | :--- | :--- |
| **Main-Store-Checkout** | `scen-123` | `https://shop.com/checkout` |
| **User-Login-Flow** | `scen-456` | `https://shop.com/login` |
| **Inventory-Search** | `scen-789` | `https://shop.com/search` |
| **Mobile-App-Smoke** | `scen-012` | `https://shop.com/mobile` |

Which one would you like to run or inspect?
```

**Prompt:** 
```
Run the checkout scenario with 200 users for 5 minutes.
```

**Response:** 
```
Triggering load test scenario `scen-123`… 

I've started the run with **200 virtual users** for a duration of **5 minutes**. 

**New Run ID:** `run-98765` 

I'll monitor the progress and notify you once the performance metrics are available.
```

**Prompt:** 
```
What were the stats for the last run?
```

**Response:** 
```
Retrieving metrics for run ID `run-13579`… 

**Performance Summary:**
* **Peak VUs:** 250
* **Avg Throughput:** 45 requests/sec
* **P95 Duration:** 1.2s

No major bottlenecks were detected during this run. Would you like to see the raw metrics breakdown?
```

## Capabilities

### Trigger new load tests
Start a test run with specific virtual user counts and durations.

### Kill active test runs
Stop a running test immediately to save resources or halt errors.

### Pull raw performance stats
Get detailed metrics like throughput and P95 latency for specific runs.

### List all test scenarios
View every load test configuration saved in your LoadNinja account.

### Check data center locations
See which physical regions are available for global load injection.

### View completion statuses
Check the real-time status of your past and active test runs.

### Audit account limits
Verify your subscription details and virtual user constraints.

## Use Cases

### Simulating a high-traffic checkout
A performance engineer asks their agent to run the 'Checkout' scenario with 500 VUs for 10 minutes and then report the P95 latency.

### Verifying a QA regression
A QA lead asks the agent to list all test runs and check the status of the one from an hour ago to see if the fix held up.

### Planning a global rollout
A DevOps pro asks the agent to list all available data center locations to ensure they have coverage in Europe and Asia before a launch.

### Emergency test termination
A test goes wrong and starts spiking CPU. The user tells the agent to stop the currently running test immediately to protect the staging server.

## Benefits

- Stop jumping between tabs by triggering new load tests directly through your AI client using run_scenario.
- Get instant visibility into bottlenecks by pulling raw performance stats with get_test_run_stats during a crunch.
- Manage your infrastructure better by checking available data center locations with list_locations before a big launch.
- Save time on reporting by having your agent summarize completed test runs using get_test_run.
- Keep your environment stable by quickly killing runaway tests with stop_test_run without searching for a dashboard.
- Audit your subscription capacity easily by checking your VU limits with get_account.

## How It Works

The bottom line is it gives you a conversational interface for your entire LoadNinja performance testing suite.

1. Connect your LoadNinja API key to the Connector.
2. Tell your agent to list, run, or monitor specific scenarios.
3. Get instant feedback on test results and infrastructure limits.

## Frequently Asked Questions

**How does LoadNinja MCP help with load testing?**
It gives you a conversational interface to manage your performance tests. You can trigger runs, check stats, and audit your infrastructure directly through your AI client without leaving your workspace.

**Can I stop a test if it crashes the site?**
Yes. You can tell your agent to immediately terminate an active load test. This is helpful for stopping runaway tests that might be impacting your staging environment.

**How do I see my VU limits?**
You can simply ask your agent to check your account details. It will pull your subscription info and show you exactly how many virtual users you have available for load injection.

**Can I check which countries are supported?**
Yes. Your agent can list all the physical data center locations available on your plan, helping you plan global load distribution for your next launch.

**Can I get raw metrics for a specific run?**
Absolutely. You can ask for the raw performance statistics of any completed run, and the agent will provide details on throughput, peak users, and transaction times.

**Does this work with Claude and Cursor?**
Yes. This Connector is designed to work with any MCP-compatible client, including Claude, Cursor, Windsurf, and VS Code.

**What is unique about LoadNinja's approach to load testing?**
Unlike protocol-based tools, LoadNinja uses actual real browsers at scale. Your agent can trigger scenarios that execute scripts inside native browsers, ensuring your performance metrics accurately reflect the true end-user experience including client-side execution.

**Can I stop a running test through my agent if I see performance degradation?**
Yes. Use the `stop_test_run` tool with the active Run ID. Your agent will dispatch an immediate halt command to LoadNinja, stopping all physical load processes and allowing you to analyze the metrics captured up to that point.

**How do I check my account's virtual user (VU) limits?**
The `get_account` tool retrieves your subscription details and strict runtime VU limits. Your agent will report your current usage caps, ensuring you stay within your performance testing budget.