feat(ai): Implement Gemini provider with streaming and tool support

- Added GeminiLLM provider implementation with GoogleGenerativeAI SDK
- Supports streaming with text/thinking content and completion signals
- Handles Gemini's parts-based content system (text, thought, functionCall)
- Implements tool/function calling with proper format conversion
- Maps between unified types and Gemini-specific formats (model vs assistant role)
- Added test example matching other provider patterns
- Fixed typo in AssistantMessage type (stopResaon -> stopReason) across all providers

packages/ai/src/providers/openai-completions.ts

@@ -163,7 +163,7 @@ export class OpenAICompletionsLLM implements LLM<OpenAICompletionsLLMOptions> {
 				toolCalls: toolCalls.length > 0 ? toolCalls : undefined,
 				model: this.model,
 				usage,
-				stopResaon: this.mapStopReason(finishReason),
+				stopReason: this.mapStopReason(finishReason),
 			};
 		} catch (error) {
 			return {
@@ -175,7 +175,7 @@ export class OpenAICompletionsLLM implements LLM<OpenAICompletionsLLMOptions> {
 					cacheRead: 0,
 					cacheWrite: 0,
 				},
-				stopResaon: "error",
+				stopReason: "error",
 				error: error instanceof Error ? error.message : String(error),
 			};
 		}