> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.qa.army/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.qa.army/_mcp/server.

# Run Configuration

> Plan environments, ordering, viewport, location, and shared setup for a test run.

> **Warning**
>
> Direct Web Runs are available. Review the saved Test and Project target before
> starting execution. Native Android, iOS, and Desktop execution are not generally
> available.

## Choose the target

Confirm the approved Web Project target before starting a Run. Record the
required viewport and location in the saved Test where supported. For Android or iOS, record the app build and
device needs. For Desktop, record the application and operating-system needs.

## Keep ordering explicit

Tests contain an ordered list of steps. If one journey depends on another,
prefer a clear shared setup within its Test Group over an undocumented timing
assumption. The CLI and local MCP can dispatch an explicit selection of saved
Tests from a Group; each selected Test produces its own Run. Dispatch is not a
passing group result, and Tests must not depend on another selected Run having
finished. See [Run selected Tests from a Group](/interfaces/cli#run-selected-tests-from-a-group).

## Make the result decidable

Every required assertion needs an observable expected outcome. Missing access,
setup, or required evidence should produce `ERROR`, not `FAILED`.