What Is ASO App Store Screenshots?
ASO App Store Screenshots is a Claude Code skill built by adamlyttleapps that automates generation of high-converting iOS App Store screenshots. It scans your iOS app's codebase to pull 3-5 core user benefits, rates existing simulator screenshots, pairs them optimally, and outputs polished PNGs at exact Apple dimensions. ASO App Store Screenshots is one of the best Claude Code Skills for iOS developers handling App Store Optimization (ASO). With 405 GitHub stars as of February 2026, it uses Python's Pillow for compositing and Gemini MCP for AI polish, saving manual design time from hours to minutes per screenshot set.
Quick Overview
| Attribute | Details |
|---|---|
| Type | Claude Code Skills |
| Best For | iOS developers |
| Language/Stack | Python, Claude Code, Gemini MCP |
| License | MIT |
| GitHub Stars | 405 as of Feb 2026 |
| Pricing | Open-Source |
| Last Release | N/A — latest commit 87e1c80 on recent date |
Who Should Use ASO App Store Screenshots?
- Indie iOS hackers launching MVPs who need ASO screenshots without hiring designers, as it pulls benefits directly from Swift/Objective-C code.
- Small mobile teams iterating on App Store listings weekly, where pairing simulator shots with codebase insights cuts A/B test cycles.
- Claude Code power users already in Anthropic's ecosystem, integrating ASO into their AI-driven dev workflow.
Not ideal for:
- Android developers, since it targets iOS device frames and Apple screenshot specs only.
- Teams without macOS, due to SF Pro Display Black font dependency from Apple developer fonts.
- Non-Claude Code users lacking MCP server setup for Gemini enhancement.
Key Features of ASO App Store Screenshots
- Benefit Discovery — Parses app codebase files like Info.plist, Swift sources, and README to extract 3-5 quantifiable benefits, such as "reduces photo editing time by 70%" from code comments or strings.
- Screenshot Rating and Pairing — Scans simulator PNGs in your project, scores them on clarity/relevance (0-10 scale), and matches highest-rated shots to benefits using keyword similarity via Claude's analysis.
- Deterministic Scaffolding —
compose.pyscript positions text exactly with Pillow: headline in SF Pro Display Black at 60pt, benefit slug below, device mockup frame at 1242x2688px for iPhone 15 Pro Max. - AI Enhancement Pipeline — Pipes scaffolds to Nano Banana Pro model via @houtini/gemini-mcp, adding photorealistic shadows, breakout UI elements, and gradient backgrounds in under 30s per image.
- Versioning and Selection — Outputs v1.png to v3.png variants per scaffold; user selects via Claude interactive prompt, auto-resizes to App Store specs (e.g., 1242x2688 for 6.7" display).
- Showcase Preview —
showcase.pycomposites all final screenshots side-by-side into a single PNG for quick App Store listing review. - Resume-Friendly State — Saves progress in Claude Code's memory, allowing pause/resume across sessions without re-analyzing codebase.
ASO App Store Screenshots vs Alternatives
| Tool | Best For | Key Differentiator | Pricing |
|---|---|---|---|
| ASO App Store Screenshots | iOS devs in Claude Code | Codebase-driven benefit extraction + deterministic scaffold | Open-Source |
| Claude Code Canvas | Visual prototyping in Claude | Canvas-based UI mocking, no ASO focus | Open-Source |
| MakeAppIcon | Icon/screenshot batch resizing | Simple drag-drop, lacks AI benefit analysis | Freemium |
| AppStore Screenshot Generator | Template-based mocks | Pre-built frames, no codebase integration | Paid |
Claude Code Canvas excels for interactive UI wireframing during iOS dev but skips ASO-specific benefit pairing—pair it with ASO App Store Screenshots for full listing optimization. MakeAppIcon handles bulk resizing across devices faster (under 10s for 10 shots) but requires manual benefit text input, missing automated codebase scan. AppStore Screenshot Generator offers 50+ templates with one-click export yet produces generic results without your app's real simulator shots or AI polish.
How ASO App Store Screenshots Works
ASO App Store Screenshots runs as a Claude Code slash command (/aso-appstore-screenshots), triggering a four-phase pipeline: discovery, pairing, generation, showcase. Core abstraction is a scaffold-enhance loop: compose.py builds pixel-perfect bases using Pillow's ImageDraw for text at fixed coordinates (e.g., title at (100, 200), benefit at (100, 800)), composites your PNG simulator shot at (200, 500)-(1042, 2200), and adds iPhone bezel SVG overlay. This deterministic step ensures layout consistency across 5-10 screenshots, avoiding AI hallucinations in positioning.
Enhancement routes scaffolds to Gemini MCP server, where Nano Banana Pro model applies diffusion-based refinements: device reflections at 20% opacity, text glow via kernel blur (5px radius), and contextual elements like floating notifications matching your app's UI from codebase screenshots. Output lands in screenshots/01-benefit-slug/ with intermediates for inspection. State persists via Claude's conversation memory JSON, tracking selected versions.
# From iOS project dir with Claude Code active
claude install-skill github.com/adamlyttleapps/claude-skill-aso-appstore-screenshots
pip install Pillow
# Install SF Pro: download from developer.apple.com/fonts
npm install -g @houtini/gemini-mcp
# Add to ~/.claude/settings.json: {"mcp_servers": ["gemini-mcp"]}
/aso-appstore-screenshots
This installs the skill, deps, and MCP; running the command prompts codebase scan (e.g., "Found benefits: fast-edits, ai-filters"), then screenshot upload/review. Expect 2-5min for full set on M1 Mac, with interactive Claude prompts for approvals at each phase.
Pros and Cons of ASO App Store Screenshots
Pros:
- Automates benefit extraction from real codebase, yielding authentic copy like "Batch process 100+ photos in 5s" pulled from your Swift funcs—beats manual brainstorming.
- Two-stage pipeline delivers consistent 1242x2688px outputs compliant with App Store Review Guidelines 2.3.10, zero rejections from dimension mismatches.
- Integrates natively with Claude Code workflow; resume mid-process via memory saves 30min+ on interruptions.
- MIT-licensed with full source; fork to tweak scaffolds for iPadOS or custom fonts in <50 LOC.
- Generates 3 variants per shot, enabling A/B testing with 15% CTR lift potential per ASO benchmarks.
- Lightweight deps: Pillow (15MB), no heavy ML libs on your machine—offloads to Gemini MCP.
Cons:
- Requires macOS for SF Pro font path
/Library/Fonts/SF-Pro-Display-Black.otf; Linux/Windows users need font hacks. - Gemini MCP setup adds npm global install and config JSON edit, barrier for non-Node devs (5min overhead).
- iOS-only: hardcoded iPhone frames, no Android Play Store support without
generate_frame.pymods. - Interactive Claude prompts can loop on edge cases like blurry simulator shots, needing manual cleanup.
- No bulk mode for 100+ screenshots; processes 5-10 per run, suited for single-app updates not portfolios.
Getting Started with ASO App Store Screenshots
Prep your iOS project dir with simulator PNGs (e.g., Screenshots/Home.png) and Xcode build artifacts. Install via Claude Code slash command, Python Pillow, SF Pro font, and Gemini MCP as shown earlier.
# In project root
claude install-skill github.com/adamlyttleapps/claude-skill-aso-appstore-screenshots
pip install Pillow
# Download SF-Pro-Display-Black.otf to /Library/Fonts/
npm i -g @houtini/gemini-mcp
# Edit ~/.claude/settings.json
/aso-appstore-screenshots
Post-run, Claude analyzes files (scans .swift, .m, plist for strings like "userBenefit"), lists benefits, rates your PNGs (e.g., "Home.png: 8/10 clarity"), pairs them, generates scaffolds in screenshots/, enhances via Gemini (v1-v3), prompts selection, resizes finals to screenshots/final/, and builds showcase.png. Tweak compose.py widths (e.g., descriptor max 800px) if text wraps oddly; rerun phases individually via Claude memory.
Verdict
ASO App Store Screenshots is the strongest Claude Code skill for iOS developers automating ASO screenshots when you have simulator shots and want codebase-tied benefits. Its scaffold + Gemini pipeline ensures pixel-perfect, conversion-focused outputs in minutes, outpacing manual Figma work. Caveat: MCP setup required. Install if you're in Claude ecosystem for 2x faster Store listing iterations.



