Skip to main content
Use these as direct prompts to your AI agent. Each example is written in plain language first, then shows the FlowDeck commands the agent should execute.
Agents may resolve workspace, scheme, and simulator values differently by project. That is expected. The requirement is to use FlowDeck commands for the same workflow and report the exact commands executed.
--simulator is optional in many workflows. Use it only when you need explicit targeting (for example, multiple simulators are booted).
For UI automation framework docs:

​
Build the app

Tell the agent: “Build my app with the current FlowDeck project settings.” Agent should run: 
flowdeck build

​
Run and watch logs

Tell the agent: “Run my app and stream logs.” Agent should run: 
flowdeck run --log

​
Launch a simulator first

Tell the agent: “Boot the first available simulator and open the Simulator app.” Agent should run: 
flowdeck simulator list --json
flowdeck simulator boot "<simulator-from-list>"
flowdeck simulator open

​
UI login smoke test

Tell the agent: “Start UI automation on the booted simulator, fill login, tap Login, wait for Home, and save a screenshot.” Agent should run: 
flowdeck ui simulator session start
flowdeck ui simulator tap "emailField" --by-id
flowdeck ui simulator type "user@example.com"
flowdeck ui simulator tap "passwordField" --by-id
flowdeck ui simulator type "secret123" --mask
flowdeck ui simulator tap "loginButton" --by-id
flowdeck ui simulator wait "Home" --timeout 20
flowdeck ui simulator assert visible "Home"
flowdeck ui simulator screen --output /tmp/login-result.png
flowdeck ui simulator session stop
Tell the agent: “Open this deep link in the simulator: myapp://settings/profile” Agent should run: 
flowdeck ui simulator open-url "myapp://settings/profile"

​
Run tests

Tell the agent: “Run tests with current FlowDeck project settings.” Agent should run: 
flowdeck test

​
Get machine-readable output

Tell the agent: “Run build and run commands and return JSON output only.” Agent should run: 
flowdeck context --json
flowdeck build --json
flowdeck run --json

​
Build a SwiftUI view from a mockup image (tangible)

Tell the agent: “I attached a mockup image. Implement a SwiftUI view to match it, run the app, and give me a current screenshot so I can compare.” Agent should run: 
flowdeck run --log
flowdeck ui simulator session start
flowdeck ui simulator screen --output /tmp/current-ui.png
flowdeck ui simulator session stop
Then the agent should iterate code changes and repeat run + screen until the visual match is acceptable.

​
Build a SwiftUI login screen from an image and verify IDs

Tell the agent: “Use this login mockup image, implement the SwiftUI screen, and verify emailField, passwordField, and loginButton accessibility IDs exist.” Agent should run: 
flowdeck run
flowdeck ui simulator session start
flowdeck ui simulator find "emailField" --by-id --json
flowdeck ui simulator find "passwordField" --by-id --json
flowdeck ui simulator find "loginButton" --by-id --json
flowdeck ui simulator screen --tree --json
flowdeck ui simulator session stop