feat(#131): add OpenTelemetry tracing infrastructure

Implement comprehensive distributed tracing for HTTP requests and LLM
operations using OpenTelemetry with GenAI semantic conventions.

Features:
- TelemetryService: SDK initialization with OTLP HTTP exporter
- TelemetryInterceptor: Automatic HTTP request spans
- @TraceLlmCall decorator: LLM operation tracing
- GenAI semantic conventions for model/token tracking
- Graceful degradation when tracing disabled

Instrumented:
- All HTTP requests (automatic spans)
- OllamaProvider chat/chatStream/embed operations
- Token counts, model names, durations

Environment:
- OTEL_ENABLED (default: true)
- OTEL_SERVICE_NAME (default: mosaic-api)
- OTEL_EXPORTER_OTLP_ENDPOINT (default: localhost:4318)

Tests: 23 passing with full coverage

Fixes #131

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

apps/api/src/llm/providers/ollama.provider.ts

@@ -6,6 +6,8 @@ import type {
   LlmProviderHealthStatus,
 } from "./llm-provider.interface";
 import type { ChatRequestDto, ChatResponseDto, EmbedRequestDto, EmbedResponseDto } from "../dto";
+import { TraceLlmCall, createLlmSpan } from "../../telemetry";
+import { SpanStatusCode } from "@opentelemetry/api";
 
 /**
  * Configuration for Ollama LLM provider.
@@ -137,6 +139,7 @@ export class OllamaProvider implements LlmProviderInterface {
    * @returns Complete chat response
    * @throws {Error} If the request fails
    */
+  @TraceLlmCall({ system: "ollama", operation: "chat" })
   async chat(request: ChatRequestDto): Promise<ChatResponseDto> {
     try {
       const messages = this.buildMessages(request);
@@ -176,6 +179,8 @@ export class OllamaProvider implements LlmProviderInterface {
    * @throws {Error} If the request fails
    */
   async *chatStream(request: ChatRequestDto): AsyncGenerator<ChatResponseDto> {
+    const span = createLlmSpan("ollama", "chat.stream", request.model);
+
     try {
       const messages = this.buildMessages(request);
       const options = this.buildChatOptions(request);
@@ -197,10 +202,21 @@ export class OllamaProvider implements LlmProviderInterface {
           done: chunk.done,
         };
       }
+
+      span.setStatus({ code: SpanStatusCode.OK });
     } catch (error: unknown) {
       const errorMessage = error instanceof Error ? error.message : String(error);
       this.logger.error(`Streaming failed: ${errorMessage}`);
+
+      span.recordException(error instanceof Error ? error : new Error(errorMessage));
+      span.setStatus({
+        code: SpanStatusCode.ERROR,
+        message: errorMessage,
+      });
+
       throw new Error(`Streaming failed: ${errorMessage}`);
+    } finally {
+      span.end();
     }
   }
 
@@ -211,6 +227,7 @@ export class OllamaProvider implements LlmProviderInterface {
    * @returns Embeddings response with vector arrays
    * @throws {Error} If the request fails
    */
+  @TraceLlmCall({ system: "ollama", operation: "embed" })
   async embed(request: EmbedRequestDto): Promise<EmbedResponseDto> {
     try {
       const response = await this.client.embed({