Android Performance Audit

Android Performance Auditing 🚀⏱️

Expert strategies for profiling and optimizing Android applications to ensure smooth, responsive, and efficient user experiences.

⚡ When to Use

• UI Lag: Fixing jank in lists or complex Compose layouts.
• Memory Issues: Debugging OutOfMemoryError or high memory growth.
• App Startup: Reducing the time from icon tap to first usable frame.
• Battery Drain: Identifying excessive network calls or background work.
• Leak Detection: Investigating persistent objects after activity destruction.

🛠️ Main Tools

1. Android Studio Profiler

• CPU Profiler: Inspect call stacks and thread activity.
• Memory Profiler: Capture heap dumps and track allocations.
• Network Profiler: Analyze traffic and payload sizes.
• Energy Profiler: Monitor battery impact.

2. LeakCanary

• Rule: ALWAYS include LeakCanary in debug builds.
• Use it to detect memory leaks automatically during development.

🎥 Compose Optimization

• [ ] Recomposition Tracking: Use the "Layout Inspector" in Android Studio to count recompositions.
• [ ] Stability Rules: Use @Stable or @Immutable on domain models to help the Compose compiler.
• [ ] Lambda Stability: Avoid pass-through lambdas that capture unstable variables; use method references where possible.
• [ ] remember & derivedStateOf: Minimize expensive calculations within the composition loop.

🏁 App Startup Optimization

• Baseline Profiles: Use Baseline Profiles to pre-compile critical code paths (AOT) to improve startup by up to 30%.
• Lazy Initialization: Use Lazy or Hilt entry points to avoid unnecessary object creation during Application.onCreate.
• App Startup Library: Utilize androidx.startup for organized component initialization.

📦 Memory Management Checklist

• [ ] Context Leaks: Ensure no long-lived references to Activity context (use ApplicationContext if possible).
• [ ] Bitmap Scaling: Never load full-resolution bitmaps into small Image views. Use Coil with size() transformations.
• [ ] Flow/Coroutine Cleanup: Use viewModelScope to ensure background tasks are canceled with the UI.

🧪 Benchmarking

• Macrobenchmark: Measure high-level user interactions (Startup, Scrolling).
• Microbenchmark: Measure isolated code loops or CPU-intensive algorithms.

🔗 Related Resources

• Android Performance Documentation
• Baseline Profiles Guide

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: prasad-vennam
• Source: prasad-vennam/Awesome-Android-AI-Agent-Skills
• License: MIT

Install and usage instructions live in the source repository linked above.