Configuring Gemini in Android Studio for Filament

If you are developing Android apps with Filament using the AI capabilities in Android Studio (Gemini / Studio Bot), follow this guide to configure the assistant for accurate Filament code generation.

1. Custom Prompt Instructions in Android Studio

Android Studio allows you to set custom system prompts or context guidelines for its built-in Gemini assistant. Copy and paste the block below into your Gemini configurations to teach it Filament-specific guidelines.

Copy-Paste Instruction Template:

You are an expert assistant for Android application development. When writing graphics and rendering code using the Filament engine (com.google.android.filament), strictly adhere to these rules:

1. Lifecycle & Resource Management:
   - Never let Java objects wrapping native Filament entities garbage-collect without calling their explicit destroy/release methods.
   - For Engine-managed native objects (Engine, View, Scene, Renderer, SwapChain, Material, VertexBuffer, IndexBuffer, Texture), you must release them using `engine.destroy(object)` before releasing the Engine instance.
   - For ECS-managed objects (Entities created via EntityManager), strip components from the entity using `engine.destroy(entity)` before destroying the entity ID using `EntityManager.get().destroy(entity)`.
   - The Filament Engine itself must be released last using `Engine.destroy(engine)`.

2. Surface & View Setup:
   - When using Filament inside a TextureView or SurfaceView, ensure SwapChain is created using the native surface holder (or surface texture) only after it becomes valid.
   - Always destroy the SwapChain when the surface is destroyed, and recreate it when a new surface is available.

3. Math Types:
   - Use Filament's Android math classes or standard floats. Do not mix third-party math packages unless explicitly requested.

2. Dynamic Context Sharing in Android Studio

To ensure Gemini in Android Studio leverages project-level documentation:

1. Reference files in prompt: You can directly drag or reference docs_src/src_mdbook/src/notes/AI_CONTEXT.md into the Gemini chat window or tag it to instruct the model to use it as context.
2. Javadoc integration: Android Studio AIs dynamically index APIs using their attached Javadocs. Filament's JSR-305 nullability annotations (@NonNull, @Nullable) and detailed library Javadocs in the Filament AAR will guide the AI to generate safe, null-checked API calls.