AutomateAI  /  Integrations
Integrations

Bring your own tools.

AutomateAI exposes its testing agents over the Model Context Protocol (MCP), so the AI tooling your team already uses can trigger runs and read results directly — no separate dashboard required.

Works with

Any MCP-compatible client

Connect locally for development, or remotely for CI/CD — both use the same set of tools.

Local — stdio

Runs alongside your IDE for day-to-day development and debugging.

Claude Desktop Cursor Windsurf VS Code Copilot

Remote — HTTP / SSE

Multi-tenant hosted endpoint with API-key auth, built for pipelines.

CI/CD pipelines Custom agents Any MCP-speaking client
Exposed tools

What your tooling can call

ToolWhat it does
list_devicesLists devices currently available on your connected farm.
run_testStarts a test run for a given prompt or saved test against web, mobile, or chatbot.
get_test_resultFetches the pass/fail verdict and step-by-step detail for a run.
get_device_screenshotPulls the current screen from a connected device mid-run.
farm_statusReports which farm nodes and devices are online.
list_promptsReturns saved test prompts available to run.
get_run_historyReturns past runs and their outcomes for a given test.