Feat/opentelemetry instrumentation

.changeset/opentelemetry-instrumentation.md

@@ -0,0 +1,7 @@
+---
+"@tanstack/ai-opentelemetry": minor
+"@tanstack/ai-openai": patch
+"@tanstack/ai-anthropic": patch
+---
+
+Add `@tanstack/ai-opentelemetry` package for OpenTelemetry instrumentation that bridges aiEventClient events to OTEL spans following GenAI Semantic Conventions. Remove console.log/error from openai and anthropic adapters for security.

packages/typescript/ai-anthropic/src/anthropic-adapter.ts

@@ -99,16 +99,6 @@ export class Anthropic extends BaseAdapter<
         this.generateId(),
       )
     } catch (error: any) {
-      console.error('[Anthropic Adapter] Error in chatStream:', {
-        message: error?.message,
-        status: error?.status,
-        statusText: error?.statusText,
-        code: error?.code,
-        type: error?.type,
-        error: error,
-        stack: error?.stack,
-      })
-
       // Emit an error chunk
       yield {
         type: 'error',
@@ -613,16 +603,6 @@ export class Anthropic extends BaseAdapter<
         }
       }
     } catch (error: any) {
-      console.error('[Anthropic Adapter] Error in processAnthropicStream:', {
-        message: error?.message,
-        status: error?.status,
-        statusText: error?.statusText,
-        code: error?.code,
-        type: error?.type,
-        error: error,
-        stack: error?.stack,
-      })
-
       yield {
         type: 'error',
         id: generateId(),

packages/typescript/ai-openai/src/openai-adapter.ts

@@ -116,10 +116,6 @@ export class OpenAI extends BaseAdapter<
         () => this.generateId(),
       )
     } catch (error: any) {
-      console.error('>>> chatStream: Fatal error during response creation <<<')
-      console.error('>>> Error message:', error?.message)
-      console.error('>>> Error stack:', error?.stack)
-      console.error('>>> Full error:', error)
       throw error
     }
   }
@@ -215,8 +211,6 @@ export class OpenAI extends BaseAdapter<
     let responseId: string | null = null
     let model: string = options.model
 
-    const eventTypeCounts = new Map<string, number>()
-
     try {
       for await (const chunk of stream) {
         chunkCount++
@@ -466,14 +460,6 @@ export class OpenAI extends BaseAdapter<
         }
       }
     } catch (error: any) {
-      console.log(
-        '[OpenAI Adapter] Stream ended with error. Event type summary:',
-        {
-          totalChunks: chunkCount,
-          eventTypes: Object.fromEntries(eventTypeCounts),
-          error: error.message,
-        },
-      )
       yield {
         type: 'error',
         id: generateId(),

packages/typescript/ai-opentelemetry/README.md

@@ -0,0 +1,122 @@
+# @tanstack/ai-opentelemetry
+
+OpenTelemetry instrumentation for TanStack AI. Automatically creates traces and spans for AI operations following [GenAI Semantic Conventions](https://opentelemetry.io/docs/specs/semconv/gen-ai/).
+
+## Installation
+
+```bash
+npm install @tanstack/ai-opentelemetry @opentelemetry/api
+```
+
+## Quick Start
+
+```typescript
+import { enableOpenTelemetry } from '@tanstack/ai-opentelemetry'
+
+// Enable instrumentation (uses global tracer provider)
+enableOpenTelemetry()
+
+// Now all TanStack AI operations will be traced
+import { chat } from '@tanstack/ai'
+import { openai } from '@tanstack/ai-openai'
+
+const stream = chat({
+  adapter: openai(),
+  model: 'gpt-4o',
+  messages: [{ role: 'user', content: 'Hello!' }],
+})
+```
+
+## Configuration
+
+```typescript
+import { enableOpenTelemetry } from '@tanstack/ai-opentelemetry'
+import { trace } from '@opentelemetry/api'
+
+enableOpenTelemetry({
+  // Provide custom tracer
+  tracer: trace.getTracer('my-app', '1.0.0'),
+
+  // Record prompt/response content (disabled by default for privacy)
+  recordContent: true,
+
+  // Record tool call arguments and results (enabled by default)
+  recordToolCalls: true,
+})
+```
+
+## Spans Created
+
+The instrumentation creates the following spans:
+
+### Chat Span
+- **Name**: `chat {model}`
+- **Kind**: CLIENT
+- **Attributes**:
+  - `gen_ai.system`: Provider name (openai, anthropic, etc.)
+  - `gen_ai.request.model`: Requested model
+  - `gen_ai.response.model`: Actual model used
+  - `gen_ai.response.finish_reasons`: How the response ended
+  - `gen_ai.usage.input_tokens`: Prompt tokens
+  - `gen_ai.usage.output_tokens`: Completion tokens
+  - `gen_ai.usage.total_tokens`: Total tokens
+
+### Stream Span
+- **Name**: `stream {model}`
+- **Kind**: INTERNAL
+- **Attributes**:
+  - `tanstack_ai.stream.total_chunks`: Number of chunks received
+  - `tanstack_ai.stream.duration_ms`: Stream duration
+
+### Tool Span
+- **Name**: `tool {toolName}`
+- **Kind**: INTERNAL
+- **Attributes**:
+  - `gen_ai.tool.name`: Tool name
+  - `gen_ai.tool.call_id`: Unique tool call ID
+  - `gen_ai.tool.duration_ms`: Execution duration
+
+## Example with Jaeger
+
+```typescript
+import { NodeSDK } from '@opentelemetry/sdk-node'
+import { JaegerExporter } from '@opentelemetry/exporter-jaeger'
+import { enableOpenTelemetry } from '@tanstack/ai-opentelemetry'
+
+// Setup OpenTelemetry SDK
+const sdk = new NodeSDK({
+  traceExporter: new JaegerExporter({
+    endpoint: 'http://localhost:14268/api/traces',
+  }),
+  serviceName: 'my-ai-app',
+})
+sdk.start()
+
+// Enable TanStack AI instrumentation
+enableOpenTelemetry()
+```
+
+## Disabling Instrumentation
+
+```typescript
+import { disableOpenTelemetry } from '@tanstack/ai-opentelemetry'
+
+disableOpenTelemetry()
+```
+
+## Manual Instrumentation
+
+For more control, you can create and manage the instrumentation instance directly:
+
+```typescript
+import { TanStackAIInstrumentation } from '@tanstack/ai-opentelemetry'
+
+const instrumentation = new TanStackAIInstrumentation({
+  recordContent: true,
+})
+
+instrumentation.enable()
+
+// Later...
+instrumentation.disable()
+```

packages/typescript/ai-opentelemetry/package.json

@@ -0,0 +1,54 @@
+{
+  "name": "@tanstack/ai-opentelemetry",
+  "version": "0.0.3",
+  "description": "OpenTelemetry instrumentation for TanStack AI",
+  "author": "",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "git+https://github.com/TanStack/ai.git",
+    "directory": "packages/typescript/ai-opentelemetry"
+  },
+  "keywords": [
+    "ai",
+    "opentelemetry",
+    "tracing",
+    "observability",
+    "tanstack",
+    "instrumentation"
+  ],
+  "type": "module",
+  "module": "./dist/esm/index.js",
+  "types": "./dist/esm/index.d.ts",
+  "exports": {
+    ".": {
+      "types": "./dist/esm/index.d.ts",
+      "import": "./dist/esm/index.js"
+    }
+  },
+  "files": [
+    "dist",
+    "src"
+  ],
+  "scripts": {
+    "build": "vite build",
+    "clean": "premove ./build ./dist",
+    "lint:fix": "eslint ./src --fix",
+    "test:build": "publint --strict",
+    "test:eslint": "eslint ./src",
+    "test:lib": "vitest",
+    "test:lib:dev": "pnpm test:lib --watch",
+    "test:types": "tsc"
+  },
+  "dependencies": {
+    "@opentelemetry/api": "^1.9.0",
+    "@opentelemetry/semantic-conventions": "^1.28.0"
+  },
+  "devDependencies": {
+    "@tanstack/ai": "workspace:*",
+    "@vitest/coverage-v8": "4.0.14"
+  },
+  "peerDependencies": {
+    "@tanstack/ai": "workspace:*"
+  }
+}

packages/typescript/ai-opentelemetry/src/index.ts

@@ -0,0 +1,10 @@
+export {
+  TanStackAIInstrumentation,
+  enableOpenTelemetry,
+  disableOpenTelemetry,
+} from './instrumentation'
+
+export {
+  GenAIAttributes,
+  type TanStackAIInstrumentationConfig,
+} from './types'