ai-docs/docs/ai/android.md
2026-02-10 15:50:59 -06:00

64 lines
1.9 KiB
Markdown

# Android AI Setup
## Setup: VS Code + GitHub Copilot Extension
### Prerequisites
- VS Code installed and up to date.
- GitHub account with Copilot Enterprise access.
### Install and Sign In (High-Level)
1. Install the GitHub Copilot extension.
2. Sign in with your GitHub account.
3. Confirm Copilot is enabled in VS Code.
4. Run a simple prompt to verify suggestions appear.
### Example Prompt (VS Code)
Open a Kotlin file and add:
Example prompt:
```text
// Create a Kotlin data class for a user profile with name and email.
```
### Verification Steps
- Open a Kotlin file and start a small function.
- Confirm inline suggestions appear.
- Open Copilot Chat and ask a short question.
## Android-Specific Guidance
- Ask for Kotlin/Java patterns with small, bounded tasks.
- Request unit tests for services and view models.
- Prefer incremental changes over large rewrites.
### Example Request
Example prompt:
```text
Refactor this repository to reduce duplication. Keep the public API the same and list tests to update.
```
## MCP For Android (High-Level)
MCP tools can automate Android tasks like builds, tests, and diagnostics. The exact tool depends on your team setup.
### What To Add Here
When you identify the approved Android MCP tool(s), add:
- Install steps
- VS Code configuration
- Example prompts
### Starter Prompts
Example prompts:
```text
Create a Kotlin data class and mapper for this API response.
Write unit tests for this repository using JUnit.
Refactor this file to reduce duplication without changing behavior.
```
## Android Troubleshooting
- If suggestions are missing, confirm sign-in and access.
- If the extension is disabled, check extension settings.
- If responses are blocked, check network or policy constraints.
### Common Setup Gaps
- Copilot access not provisioned for the GitHub account.
- VS Code extension disabled after an update.
- Multiple AI extensions competing for suggestions.