# TestMu AI MCP for AI Agents AI Agent Connect

> TestMu AI MCP lets you manage cross-browser testing and automation builds through your AI client. You can audit Selenium, Playwright, and Cypress suites, grab bug screenshots, and inspect session logs without leaving your IDE. It turns your AI agent into a QA manager that can pull telemetry and visual artifacts from the TestMu AI grid instantly.

## Overview
- **Category:** ship-it
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_OF39VBbDG3Hh15In0o7s95aFeox1GIJ02zw4AH1a/ai-agent-connect
- **Tags:** cross-browser-testing, automation-testing, selenium, playwright, visual-testing, ci-cd

## Description

TestMu AI MCP lets you connect your account to any AI agent and take full control of your cross-browser testing and automation orchestration through natural conversation. Stop jumping between your IDE and a browser dashboard just to figure out why a test failed in the CI/CD pipeline. You no longer have to open a browser, log into a dashboard, find the right test ID, and then scroll through logs to find a single console error. This Connector connects your TestMu AI account directly to your agent, letting you handle the heavy lifting of QA orchestration through a simple chat. Instead of hunting for specific session IDs or digging through nested folders of logs, you just ask your agent to find the latest failed build or pull the screenshots from a specific test run. It handles the tedious parts of testing, like checking which macOS or Android versions are currently active in the grid or grabbing the exact W3C WebDriver logs you need to squash a bug. You can see the big picture of your automation health while still having the ability to drill down into the exact console errors of a single browser session. It's a massive time saver for anyone who needs to move fast on UI regressions and wants to stay in their flow. You'll find it's one of the most practical ways to use the Vinkius catalog to bridge the gap between your code and your testing infrastructure.

## Tools

### list_builds
List all test automation builds on the TestMu AI Cloud. This helps you see your entire automation history at a glance.

### list_sessions
List recent automated test sessions including OS, browser version, and duration. You can quickly see what's running on the grid right now.

### get_session
Get deep context for a specific test session, including console errors and network configurations. This is the quickest way to find out why a specific UI interaction failed.

### get_session_logs
Extract precise Selenium or Appium command logs for a specific test. Pull these logs to see the exact steps the driver took before a crash.

### get_screenshots
Fetch visual bug screenshots captured during a test session. Use this to see the exact UI state at the moment of failure without leaving your agent.

### get_video
Retrieve the full video recording URL for a test execution. Watch the full playback to see how a user journey unfolded in a real browser.

### get_build
Get exact telemetry for a specific TestMu AI build. Use this to check pass rates and execution times for a specific run.

### list_tunnels
List all active secure UnderTest tunnels. Use this to verify your local environment is correctly exposed for testing.

### list_platforms
List all supported OS and browser capabilities available in the testing grid. Check this to see which mobile or desktop versions are currently live.

### get_build_sessions
Extract all explicit test sessions associated with a specific build. Use this to isolate deployment regressions by filtering sessions under a specific CI/CD build.

## Prompt Examples

**Prompt:** 
```
List the last 5 automation builds in my TestMu AI account
```

**Response:** 
```
I've retrieved the latest 5 builds from your TestMu AI account:

| Build Name | Status | Sessions |
| :--- | :--- | :--- |
| **Prod-Regression-987** | ✅ Passed | 45 |
| **Staging-UI-Check-986** | ❌ Failed | 12 |
| **Critical-Hotfix-985** | ✅ Passed | 3 |
| **Nightly-Smoke-984** | ✅ Passed | 20 |
| **Dev-Integration-983** | ❌ Failed | 8 |

Would you like to see the test sessions for the failed staging build?
```

**Prompt:** 
```
Show me the logs and screenshots for session ID 'abc-123-xyz'
```

**Response:** 
```
Retrieving data for session **abc-123-xyz**...

I've fetched 3 failure screenshots and the full Selenium command logs. 

**Console Summary:**
- `NoSuchElementException` at login flow.
- `TimeoutException` on submit button.

**Visual Artifacts:**
- [Screenshot 1](https://example.com/screenshot1.png)
- [Screenshot 2](https://example.com/screenshot2.png)

Would you like the video recording link as well?
```

**Prompt:** 
```
Which OS and browser versions are currently supported in the grid?
```

**Response:** 
```
I've queried the platform capabilities. Currently supported:

- **Windows 11**: Chrome 120+, Firefox 118+
- **macOS Sonoma**: Safari 17, Chrome 120+
- **Android**: Latest versions (Samsung S23, Pixel 8)
- **iOS**: Latest versions (iPhone 14, 13)

I can provide the full JSON matrix if you need specific sub-versions.
```

## Capabilities

### Audit automation builds
Get high-level telemetry for Selenium, Playwright, or Cypress suites directly from your agent.

### Pull bug screenshots
Fetch absolute URLs for visual artifacts captured during a test session.

### Inspect session logs
Extract command-level logs including W3C WebDriver protocols or Playwright RPC calls.

### Check grid capabilities
Query the grid for supported macOS, Windows, iOS, and Android versions.

### Monitor secure tunnels
List active secure tunnels for localhost testing to ensure your environment is exposed.

### Analyze build sessions
Filter and retrieve all child test sessions nested under specific CI/CD builds.

## Use Cases

### Debugging a flaky UI test
A dev asks the agent to find the last 5 failed sessions for a specific build. The agent uses get_build_sessions and then pulls screenshots with get_screenshots to identify a CSS layout shift.

### QA Health Check
A QA engineer asks for a summary of the latest Selenium builds. The agent uses list_builds and get_build to report on pass rates and execution times across the fleet.

### Mobile Capability Audit
A dev needs to know if they can test on a specific iOS version. They ask the agent to check list_platforms to see the current grid availability.

### Log Extraction for a Crash
A dev finds a crash in the CI. They ask the agent to pull the get_session_logs for that specific test ID to see the exact Playwright RPC calls that failed.

## Benefits

- Stop manual dashboard hunting. Use list_builds to see your whole automation history in one place without clicking through pages.
- Faster bug reproduction. Grab visual evidence with get_screenshots and get_video to see exactly what went wrong in a real browser.
- Deep-dive debugging. Use get_session and get_session_logs to see console errors and W3C logs without switching contexts.
- Real-time grid awareness. Check list_platforms to see exactly which macOS, Windows, or Android versions are ready for your next test.
- Simplified CI/CD mapping. Use get_build_sessions to isolate deployment regressions by filtering sessions under specific builds.
- Localhost testing visibility. Monitor your active secure tunnels with list_tunnels to ensure your local environment is correctly exposed.

## How It Works

The bottom line is you get a direct line to your testing grid without the tab-switching.

1. Subscribe to the TestMu AI MCP on Vinkius.
2. Enter your TestMu AI Username and Access Key in your client settings.
3. Ask your agent to list builds, pull logs, or check grid status.

## Frequently Asked Questions

**How can I use TestMu AI MCP to see my test results?**
You can ask your agent to list recent builds or specific sessions. It will pull the status, duration, and pass/fail rates directly into your chat window.

**Can TestMu AI MCP help me find bug screenshots?**
Yes, it can fetch the absolute URLs for screenshots captured during a test. You can ask for screenshots from a specific session ID to see exactly what went wrong.

**Does TestMu AI MCP support Playwright and Selenium?**
It supports both. You can use it to pull command logs, telemetry, and visual artifacts for your Selenium, Playwright, and Cypress automation suites.

**How do I check available mobile devices with TestMu AI MCP?**
Just ask your agent to list the supported platforms. It will return the current list of macOS, Windows, iOS, and Android versions available in the grid.

**Can I use TestMu AI MCP to debug CI/CD failures?**
Absolutely. You can ask the agent to find failed builds from your CI/CD history and pull the specific logs and screenshots needed to diagnose the regression.

**What kind of logs can I get from TestMu AI MCP?**
You can extract precise Selenium command logs or Playwright RPC calls. This includes execution timestamps and console errors to help you find the root cause of a failure.

**Can I see the exact console errors from a failed browser test through my agent?**
Yes. Use the `get_session_details` tool with a specific Session ID. Your agent will retrieve full metadata including console logs, network configurations, and execution artifacts, helping you identify the root cause of the failure instantly.

**How do I retrieve the video recording of a specific test execution?**
Ask your agent to `get_session_video` for a targeted Session ID. Your agent will return the absolute MP4 URL to stream or download the complete browser session recording, which is essential for debugging flakey UI bugs.

**Can my agent check if our secure localhost tunnel is active?**
Absolutely. Use the `list_active_tunnels` tool to identify dedicated private connection endpoints. Your agent will report which TestMu AI tunnels are currently running, ensuring you can test local environments from the cloud.