Xcode Dev Loop Skill | Agent Skills

Xcode Dev Loop

The command-line loop for an Xcode project: discover, build in the background, read the results correctly, launch on a simulator, then capture screens.

This loop runs and inspects an Xcode project. It does not author Swift source or tests. Send that work to app:swift-testing, then come back here to run it.

1. Discover Before You Guess

Find the container first. List the checkout for a .xcworkspace or a .xcodeproj.
List the schemes of a workspace with xcodebuild -list -workspace <name>.xcworkspace.
List the targets, build configurations, and schemes of a project with xcodebuild -list -project <name>.xcodeproj. Use this form when you need a target name.
Build and test with the same container flag you listed with, and always name the scheme: xcodebuild test -workspace <name>.xcworkspace -scheme <Scheme>. The test action fails without a scheme.
List simulator UDIDs with xcrun simctl list devices available -j. Resolve the UDID once, then reuse it.
Never address a simulator by name. Duplicate names ("iPhone 17 Pro Max") cause "No booted simulator named …" failures. Only the listing output separates two devices that share a name, so give the listing command beside the UDID whenever you replace a name.
When a scheme needs a runtime that the default Xcode does not have (for example a watchOS beta), export DEVELOPER_DIR=/Applications/Xcode-<version>.app/Contents/Developer for the whole session. Every xcodebuild and xcrun call must select the same Xcode, not only the first call.

2. Build and Test in the Background

Never let xcodebuild hold the foreground. A clean build or a test run routinely exceeds the 600 s command timeout. A command killed at the timeout leaves no verdict.
Start the run in the background. Redirect its output to a log file in the scratchpad. Keep a handle whose exit status you can read later.
Prefer the harness's background facility (Bash with run_in_background: true) and the task handle it returns.
When only a plain shell is available, make the run record its own exit status. A later shell cannot wait for another shell's child, so wait for that status file instead. Give every attempt its own log and status path; a status file left by an earlier attempt ends the wait at once and reports the wrong result:
```
( xcodebuild ... > "$LOG" 2>&1; echo $? > "$LOG.status" ) &   # start it
until [ -f "$LOG.status" ]; do sleep 5; done                  # wait, in a later call
```
Use CODE_SIGNING_ALLOWED=NO for simulator builds.
Pin the destination as platform=iOS Simulator,id=<udid>.
When only one target matters, scope the test run with -only-testing:<TestTarget>.
Never wait with a chained foreground sleep; the harness blocks it.
Wait on the handle you kept with the harness's blocking wait (Monitor, or the task-output wait), or wait for the .status file with the loop above. Read the exit status first. Read the log's terminal marker second. Do not wait by grep alone: the loop keeps polling when xcodebuild exits before it writes a marker.

3. Read Results Correctly (Swift Testing Is Not XCTest)

Swift Testing does NOT emit XCTest's Test Case ... passed lines.
Executed 0 tests, with 0 failures is the empty XCTest summary. It proves nothing.
The real verdict is ✔ Test run with N tests in M suites passed plus ** TEST SUCCEEDED **. Check that N is greater than zero. A pattern that also matches zero proves nothing.
Grep once with the full set of terminal markers: \*\* BUILD SUCCEEDED \*\*, \*\* BUILD FAILED \*\*, \*\* TEST SUCCEEDED \*\*, \*\* TEST FAILED \*\*, Testing failed:, and ^error:. Judge the marker together with the exit status.
To confirm which suites ran, match both spellings: XCTest prints Test Suite '.*' started, and Swift Testing prints a Suite .* started record. Match them case-sensitively as printed.
Always pass -resultBundlePath, and give every attempt its own path. xcodebuild fails when the path already exists.
On failure, read xcrun xcresulttool get test-results summary --path <bundle>. It gives the structured crash reason, such as Test crashed with signal abrt. The log may repeat it as an underlying error, but the bundle is the reliable source. Do not re-run blind.
Treat The test runner hung before establishing connection and diagnostic-collection timeouts as a suspected environmental fault, not a proven diagnosis. Shut down only the simulator you pinned: xcrun simctl shutdown <udid>. Then retry once with a new result-bundle path. Use xcrun simctl shutdown all only when the user says the whole device set is disposable.
Pass TEST_RUNNER_<NAME>=<value> on the xcodebuild test command. Xcode strips the prefix and gives the variable to the test runner process. A UI test target's app under test does not receive it; set XCUIApplication.launchEnvironment for that case.
test-without-building reuses the product built earlier. It can report a success that excludes your latest source change, so rebuild with xcodebuild test before you trust a pass.

4. Install, Launch, and Drive

Install the build you just made before you launch it. Never launch whatever is already on the device: it can come from another branch.
Boot the pinned device before you install: xcrun simctl bootstatus <udid> -b boots it when it is shut down and waits for the boot to finish. simctl install needs a booted device. Do not call xcrun simctl boot first: it fails on a device that already runs.
Know the product path before you install. Pass -derivedDataPath <dir> to the build so the path is predictable, or read BUILT_PRODUCTS_DIR and FULL_PRODUCT_NAME from xcodebuild -showBuildSettings. Repeat every flag the build used on that call, including -derivedDataPath, the scheme, and the destination. Different flags report a different product path, so you would install an app from another build.
Read the bundle id from that build, not from memory: /usr/libexec/PlistBuddy -c 'Print :CFBundleIdentifier' <app-path>/Info.plist.
Prefer the simulator MCP. Call its install action with the path of the new .app, then its launch action with the bundle id. Pass the pinned UDID to both actions; their device argument is optional and defaults elsewhere.
An MCP server that is already running does not inherit a DEVELOPER_DIR you export in a shell. When the run needs a beta Xcode, use the xcrun commands with that DEVELOPER_DIR instead of the MCP.
When the MCP launch fails with Failed to spawn xcrun (via disclaimer), fall back to xcrun simctl install <udid> <app-path>, then xcrun simctl launch <udid> <bundle-id>.
Reach screens by deep link with xcrun simctl openurl <udid> <scheme>://<path>, or by a DEBUG scenario or seed mechanism when the app has one. Do not use blind coordinate taps.
Read the accessibility hierarchy before you tap. Coordinates are the last resort.

5. Screenshots

Pick ONE capture mechanism per run. Use xcrun simctl io <udid> screenshot <path>.png when the file must persist for comparison. Use the MCP screenshot action when you only need to look now. Never use both for the same frame.
Number the files (round2/01-hub-gate.png) so rounds diff cleanly against a baseline.
A freshly erased simulator posts a system banner soon after boot. Wait for the boot to finish with xcrun simctl bootstatus <udid> -b. Then read the accessibility hierarchy and confirm that no system alert covers the view. Capture only after that check passes; a fixed short delay is not evidence.

6. Cadence

Compile after the domain layer, then again after the first view. Do not compile once at the end.
Run the test suite once, early, before you declare a change done.

Related Skills

app:debug — structured feedback loop for simulator failures, user-reported or observed
app:swift-testing — writing the tests this loop runs
app:swiftui-architecture — SwiftUI patterns for the code under test

Agent Skills: Xcode Dev Loop

Install this agent skill to your local

Skill Files