agentleFS
AlphaSquadTechAlphaSquadTech/ios-dev/SKILL.md

ios-dev

Skill2 starsChanged 9 months agoMITView on GitHub

iOS development workflow for building, running, testing, debugging, and fixing iOS apps using Xcode, Simulator, screenshots, and Maestro UI automation. Use when: building an iOS app, running on simulator, taking simulator screenshots, navigating app UI, debugging iOS issues, creating Xcode projects, writing SwiftUI or UIKit code, running xcodebuild, using xcrun simctl, or any iOS/iPhone/iPad development task.

3 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. OpenMinis/MinisSkills★ 602same name
    2. Prisma-Labs-Dev/apple-skills★ 336same name
    3. Sean-Michael/skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 0 are common, 10 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If it reports BLOCKED, fix the issues before proceeding.

    line 29
  2. Distinctive:

    If found, discover schemes:

    line 47
  3. Distinctive:

    If no project exists, create one as needed for the task.

    line 54
  4. Distinctive:

    On build failure: Read the error output, fix the source code, rebuild.

    line 88
  5. Distinctive:

    ALWAYS use the screenshot script: Never take a raw screenshot without resizing.

    line 110
  6. Very specific:

    Use descriptive names: home_screen, login_form, after_submit, error_state, detail_view.

    line 128
  7. Distinctive:

    Write a temporary Maestro flow file and run it:

    line 133
  8. Distinctive:

    Use the Read tool to view them.

    line 155
  9. Distinctive:

    When asked to build, test, or fix an iOS app, follow this exact sequence:

    line 201
  10. Distinctive:

    ALWAYS: use the Read tool to view screenshots after taking them -- this is how you see the app

    line 221
  11. Distinctive:

    NEVER: ask the user to manually interact with the simulator -- do everything programmatically

    line 222
  12. Very specific:

    ALWAYS: use CODE_SIGNING_ALLOWED=NO for simulator builds

    line 224
  13. Very specific:

    ALWAYS: use -derivedDataPath .claude-ios/build to keep build artifacts in the project

    line 225
  14. Distinctive:

    ALWAYS: run preflight checks at the start of a session

    line 226

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

iOS Development Skill

You are an expert iOS developer with full autonomous control of the Xcode build pipeline, iOS Simulator, screenshot capture, Maestro UI automation, and debug log analysis. Follow these procedures exactly.

1. Pre-flight Checks

On every invocation, run the preflight script first. Locate this skill's scripts directory:

SKILL_SCRIPTS=$(find -L .claude/skills .agents/skills -path "*/ios-dev/scripts" -type d 2>/dev/null | head -1)
bash "$SKILL_SCRIPTS/preflight.sh"

If it reports BLOCKED, fix the issues before proceeding. If Maestro is missing, the script auto-installs it. Always ensure ~/.maestro/bin is in PATH:

export PATH="$HOME/.maestro/bin:$PATH"

2. Project Discovery

Auto-detect the Xcode project:

# Prefer workspace (supports CocoaPods)
find . -maxdepth 3 -name "*.xcworkspace" -not -path "*/DerivedData/*" -not -path "*/.build/*" -not -path "*/Pods/*" 2>/dev/null | head -1

# Fall back to project file
find . -maxdepth 3 -name "*.xcodeproj" -not -path "*/DerivedData/*" -not -path "*/Pods/*" 2>/dev/null | head -1

If found, discover schemes:

xcodebuild -list -project <path> 2>/dev/null
# or
xcodebuild -list -workspace <path> 2>/dev/null

If no project exists, create one as needed for the task.

3. Simulator Management

Select a simulator

# Check if one is already booted -- use it
xcrun simctl list devices booted 2>/dev/null

# If none booted, list available iPhone simulators and pick the newest iPhone Pro:
xcrun simctl list devices available 2>/dev/null | grep "iPhone"

Selection priority: booted device > newest iPhone Pro > newest iPhone > any available.

Boot and open

xcrun simctl boot <UDID>
open -a Simulator
sleep 3  # Wait for simulator to fully boot

4. Build

xcodebuild -project <path.xcodeproj> -scheme <scheme> -sdk iphonesimulator -destination 'platform=iOS Simulator,id=<UDID>' -configuration Debug -derivedDataPath .claude-ios/build CODE_SIGNING_ALLOWED=NO build 2>&1

Note: The xcodebuild command must be on a single line when executed via Bash tool. Backslash line continuations may cause Unknown build action '' errors in agent environments.

For workspaces, use -workspace instead of -project.

On build failure: Read the error output, fix the source code, rebuild. Retry up to 3 times.

The built .app bundle will be at:

.claude-ios/build/Build/Products/Debug-iphonesimulator/<AppName>.app

5. Install & Launch

xcrun simctl install booted .claude-ios/build/Build/Products/Debug-iphonesimulator/<AppName>.app
xcrun simctl launch booted <bundle.identifier>
sleep 3  # Wait for app to load

Find the bundle identifier:

defaults read .claude-ios/build/Build/Products/Debug-iphonesimulator/<AppName>.app/Info.plist CFBundleIdentifier

6. Screenshots (CRITICAL)

ALWAYS use the screenshot script. Never take a raw screenshot without resizing.

SKILL_SCRIPTS=$(find -L .claude/skills .agents/skills -path "*/ios-dev/scripts" -type d 2>/dev/null | head -1)
bash "$SKILL_SCRIPTS/screenshot.sh" <descriptive_name>

This captures the simulator screen, resizes to max 1568px (Claude API limit), and saves to .claude-ios/screenshots/<name>.png.

After every screenshot, ALWAYS use the Read tool to view it:

Read tool -> .claude-ios/screenshots/<name>.png

This is how you see the app. Analyze what is on screen and decide next actions.

Screenshot naming convention

Use descriptive names: home_screen, login_form, after_submit, error_state, detail_view.

7. UI Navigation with Maestro

Quick single actions

Write a temporary Maestro flow file and run it:

# .claude-ios/flow.yaml
appId: <bundle.identifier>
---
- tapOn: "Button Text"
- inputText: "Hello"
- assertVisible: "Expected Result"
- takeScreenshot: .claude-ios/screenshots/after_action
export PATH="$HOME/.maestro/bin:$PATH"
MAESTRO_CLI_NO_ANALYTICS=1 maestro test .claude-ios/flow.yaml

After Maestro runs

  1. Resize any screenshots Maestro took (it does NOT auto-resize):
    sips --resampleHeightWidthMax 1568 .claude-ios/screenshots/<name>.png
  2. Use the Read tool to view them.

Key Maestro commands

  • tapOn: "text" -- tap a button/label by its text
  • tapOn: { id: "accessibilityId" } -- tap by accessibility identifier
  • inputText: "text" -- type into focused field
  • assertVisible: "text" -- verify element exists
  • scroll / swipe -- navigate lists
  • waitForAnimationToEnd -- wait after transitions
  • See references/maestro-guide.md for the full command reference.

8. Video Recording

SKILL_SCRIPTS=$(find -L .claude/skills .agents/skills -path "*/ios-dev/scripts" -type d 2>/dev/null | head -1)

# Start recording
bash "$SKILL_SCRIPTS/record.sh" start <name>

# ... perform actions ...

# Stop recording
bash "$SKILL_SCRIPTS/record.sh" stop

Videos are saved to .claude-ios/videos/.

9. Debug Logs

View recent logs

xcrun simctl spawn booted log show --last 5m --predicate 'process == "<AppName>"' --style compact

Stream logs in real-time

xcrun simctl spawn booted log stream --predicate 'process == "<AppName>"' --level debug --timeout 10

Check for crash logs

xcrun simctl spawn booted log show --last 2m --predicate 'process == "<AppName>" AND messageType == 16' --style compact

10. The Autonomous Build-Run-Verify Loop

When asked to build, test, or fix an iOS app, follow this exact sequence:

  1. Preflight -- Run preflight.sh, fix any blockers
  2. Discover -- Find or create the Xcode project
  3. Build -- xcodebuild for simulator
  4. If build fails -- Read errors, fix code, rebuild (up to 3 retries)
  5. Simulator -- Boot if needed, select best device
  6. Install -- Install .app on simulator
  7. Launch -- Start the app
  8. Screenshot -- Take screenshot via screenshot.sh, view with Read tool
  9. Verify -- Analyze the screenshot. Does the UI look correct?
  10. If UI is wrong -- Diagnose from screenshot, fix code, rebuild from step 3
  11. Navigate -- If needed, use Maestro to tap/type/swipe through the app
  12. Screenshot again -- Capture and view each significant state
  13. Debug -- If something is off, check logs for errors
  14. Iterate -- Repeat until the app works correctly

11. Rules (Non-negotiable)

  1. NEVER view a screenshot without resizing first -- always use screenshot.sh or sips --resampleHeightWidthMax 1568
  2. ALWAYS use the Read tool to view screenshots after taking them -- this is how you see the app
  3. NEVER ask the user to manually interact with the simulator -- do everything programmatically
  4. ALWAYS store artifacts in .claude-ios/screenshots/ and .claude-ios/videos/
  5. ALWAYS use CODE_SIGNING_ALLOWED=NO for simulator builds
  6. ALWAYS use -derivedDataPath .claude-ios/build to keep build artifacts in the project
  7. ALWAYS run preflight checks at the start of a session
  8. For troubleshooting, see references/troubleshooting.md

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.