Add Google Gemini API support

[Copilot]

Adds support for Google Gemini models via the official Gemini API (https://ai.google.dev/api).

Changes

Core Implementation

• Added ApiType.gemini and Provider.gemini enum cases
• Created geminiChat.swift with Gemini-specific message types:
  • GeminiMessage, GeminiUserMessage, GeminiModelMessage
  • GeminiPart supporting text, inline data, function calls/responses
  • GeminiTool for function declarations
• Created GeminiClient actor with streaming support using endpoint format: /v1beta/models/{model}:streamGenerateContent?key={apiKey}
• Updated Message enum to support both OpenAI and Gemini messages

Tests

• GeminiClientTests: Streaming, cancellation, function calling
• ParseGeminiMessageTests: Message parsing/serialization
• geminiChatController: Vapor-based mock server

Usage Example

let client = GeminiClient(
    baseURL: URL(string: "https://generativelanguage.googleapis.com")!,
    apiKey: apiKey
)

let stream = await client.generateStreamResponse(
    systemText: "You are a helpful assistant.",
    message: GeminiUserMessage(text: "Hello"),
    model: OpenAICompatibleModel(id: "gemini-pro"),
    tools: [/* GeminiTool objects */]
)

for try await message in stream.stream {
    // Handle streaming responses
}

Pattern follows existing OpenAI implementation for consistency.

Original prompt

add gemini model support: https://ai.google.dev/api

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.