feat(core): Instrument LangGraph Agent

dev-packages/node-integration-tests/package.json

@@ -31,6 +31,7 @@
     "@hono/node-server": "^1.19.4",
     "@langchain/anthropic": "^0.3.10",
     "@langchain/core": "^0.3.28",
+    "@langchain/langgraph": "^0.2.32",
     "@nestjs/common": "^11",
     "@nestjs/core": "^11",
     "@nestjs/platform-express": "^11",

dev-packages/node-integration-tests/suites/tracing/langgraph/instrument-with-pii.mjs

@@ -0,0 +1,10 @@
+import * as Sentry from '@sentry/node';
+import { loggingTransport } from '@sentry-internal/node-integration-tests';
+
+Sentry.init({
+  dsn: 'https://[email protected]/1337',
+  release: '1.0',
+  tracesSampleRate: 1.0,
+  sendDefaultPii: true,
+  transport: loggingTransport,
+});

dev-packages/node-integration-tests/suites/tracing/langgraph/instrument.mjs

@@ -0,0 +1,10 @@
+import * as Sentry from '@sentry/node';
+import { loggingTransport } from '@sentry-internal/node-integration-tests';
+
+Sentry.init({
+  dsn: 'https://[email protected]/1337',
+  release: '1.0',
+  tracesSampleRate: 1.0,
+  sendDefaultPii: false,
+  transport: loggingTransport,
+});

dev-packages/node-integration-tests/suites/tracing/langgraph/scenario-tools.mjs

@@ -0,0 +1,164 @@
+import { tool } from '@langchain/core/tools';
+import { END, MessagesAnnotation, START, StateGraph } from '@langchain/langgraph';
+import { ToolNode } from '@langchain/langgraph/prebuilt';
+import * as Sentry from '@sentry/node';
+import { z } from 'zod';
+
+async function run() {
+  await Sentry.startSpan({ op: 'function', name: 'langgraph-tools-test' }, async () => {
+    // Define tools
+    const getWeatherTool = tool(
+      async ({ city }) => {
+        return JSON.stringify({ city, temperature: 72, condition: 'sunny' });
+      },
+      {
+        name: 'get_weather',
+        description: 'Get the current weather for a given city',
+        schema: z.object({
+          city: z.string().describe('The city to get weather for'),
+        }),
+      },
+    );
+
+    const getTimeTool = tool(
+      async () => {
+        return new Date().toISOString();
+      },
+      {
+        name: 'get_time',
+        description: 'Get the current time',
+        schema: z.object({}),
+      },
+    );
+
+    const tools = [getWeatherTool, getTimeTool];
+    const toolNode = new ToolNode(tools);
+
+    // Define mock LLM function that returns without tool calls
+    const mockLlm = () => {
+      return {
+        messages: [
+          {
+            role: 'assistant',
+            content: 'Response without calling tools',
+            response_metadata: {
+              model_name: 'gpt-4-0613',
+              finish_reason: 'stop',
+              tokenUsage: {
+                promptTokens: 25,
+                completionTokens: 15,
+                totalTokens: 40,
+              },
+            },
+            tool_calls: [],
+          },
+        ],
+      };
+    };
+
+    // Routing function - check if there are tool calls
+    const shouldContinue = state => {
+      const messages = state.messages;
+      const lastMessage = messages[messages.length - 1];
+
+      // If the last message has tool_calls, route to tools, otherwise end
+      if (lastMessage.tool_calls && lastMessage.tool_calls.length > 0) {
+        return 'tools';
+      }
+      return END;
+    };
+
+    // Create graph with conditional edge to tools
+    const graph = new StateGraph(MessagesAnnotation)
+      .addNode('agent', mockLlm)
+      .addNode('tools', toolNode)
+      .addEdge(START, 'agent')
+      .addConditionalEdges('agent', shouldContinue, {
+        tools: 'tools',
+        [END]: END,
+      })
+      .addEdge('tools', 'agent')
+      .compile({ name: 'tool_agent' });
+
+    // Simple invocation - won't call tools since mockLlm returns empty tool_calls
+    await graph.invoke({
+      messages: [{ role: 'user', content: 'What is the weather?' }],
+    });
+
+    // Define mock LLM function that returns with tool calls
+    let callCount = 0;
+    const mockLlmWithTools = () => {
+      callCount++;
+
+      // First call - return tool calls
+      if (callCount === 1) {
+        return {
+          messages: [
+            {
+              role: 'assistant',
+              content: '',
+              response_metadata: {
+                model_name: 'gpt-4-0613',
+                finish_reason: 'tool_calls',
+                tokenUsage: {
+                  promptTokens: 30,
+                  completionTokens: 20,
+                  totalTokens: 50,
+                },
+              },
+              tool_calls: [
+                {
+                  name: 'get_weather',
+                  args: { city: 'San Francisco' },
+                  id: 'call_123',
+                  type: 'tool_call',
+                },
+              ],
+            },
+          ],
+        };
+      }
+
+      // Second call - return final response after tool execution
+      return {
+        messages: [
+          {
+            role: 'assistant',
+            content: 'Based on the weather data, it is sunny and 72 degrees in San Francisco.',
+            response_metadata: {
+              model_name: 'gpt-4-0613',
+              finish_reason: 'stop',
+              tokenUsage: {
+                promptTokens: 50,
+                completionTokens: 20,
+                totalTokens: 70,
+              },
+            },
+            tool_calls: [],
+          },
+        ],
+      };
+    };
+
+    // Create graph with tool calls enabled
+    const graphWithTools = new StateGraph(MessagesAnnotation)
+      .addNode('agent', mockLlmWithTools)
+      .addNode('tools', toolNode)
+      .addEdge(START, 'agent')
+      .addConditionalEdges('agent', shouldContinue, {
+        tools: 'tools',
+        [END]: END,
+      })
+      .addEdge('tools', 'agent')
+      .compile({ name: 'tool_calling_agent' });
+
+    // Invocation that actually calls tools
+    await graphWithTools.invoke({
+      messages: [{ role: 'user', content: 'What is the weather in San Francisco?' }],
+    });
+  });
+
+  await Sentry.flush(2000);
+}
+
+run();

dev-packages/node-integration-tests/suites/tracing/langgraph/scenario.mjs

@@ -0,0 +1,52 @@
+import { END, MessagesAnnotation, START, StateGraph } from '@langchain/langgraph';
+import * as Sentry from '@sentry/node';
+
+async function run() {
+  await Sentry.startSpan({ op: 'function', name: 'langgraph-test' }, async () => {
+    // Define a simple mock LLM function
+    const mockLlm = () => {
+      return {
+        messages: [
+          {
+            role: 'assistant',
+            content: 'Mock LLM response',
+            response_metadata: {
+              model_name: 'mock-model',
+              finish_reason: 'stop',
+              tokenUsage: {
+                promptTokens: 20,
+                completionTokens: 10,
+                totalTokens: 30,
+              },
+            },
+          },
+        ],
+      };
+    };
+
+    // Create and compile the graph
+    const graph = new StateGraph(MessagesAnnotation)
+      .addNode('agent', mockLlm)
+      .addEdge(START, 'agent')
+      .addEdge('agent', END)
+      .compile({ name: 'weather_assistant' });
+
+    // Test: basic invocation
+    await graph.invoke({
+      messages: [{ role: 'user', content: 'What is the weather today?' }],
+    });
+
+    // Test: invocation with multiple messages
+    await graph.invoke({
+      messages: [
+        { role: 'user', content: 'Hello' },
+        { role: 'assistant', content: 'Hi there!' },
+        { role: 'user', content: 'Tell me about the weather' },
+      ],
+    });
+  });
+
+  await Sentry.flush(2000);
+}
+
+run();