Skip to main content

​
Overview

The Anyway JS SDK provides functions and decorators to structure your traces:
  • withWorkflow - High-level operations that orchestrate multiple tasks
  • withTask - Individual units of work (like LLM calls)
  • withAgent - Agent operations with automatic agent name propagation
  • withTool - Tool executions within agents

​
Basic Setup

import { initialize, withWorkflow, withTask } from "@anyway-sh/node-server-sdk";

initialize({ appName: "my-app" });

​
withWorkflow and withTask

Use withWorkflow for top-level operations and withTask for individual steps: 
import OpenAI from "openai";
import { initialize, withWorkflow, withTask } from "@anyway-sh/node-server-sdk";

initialize({
  appName: "document-processor",
  instrumentModules: { openAI: OpenAI },
});

const client = new OpenAI();

async function summarize(text: string): Promise<string> {
  return withTask({ name: "summarize" }, async () => {
    const response = await client.chat.completions.create({
      model: "gpt-4o-mini",
      messages: [
        { role: "system", content: "Summarize the text concisely." },
        { role: "user", content: text },
      ],
    });
    return response.choices[0].message.content!;
  });
}

async function extractKeywords(text: string): Promise<string[]> {
  return withTask({ name: "extract_keywords" }, async () => {
    const response = await client.chat.completions.create({
      model: "gpt-4o-mini",
      messages: [
        { role: "system", content: "Extract 5 keywords. Return comma-separated." },
        { role: "user", content: text },
      ],
    });
    return response.choices[0].message.content!.split(", ");
  });
}

async function processDocument(document: string) {
  return withWorkflow({ name: "process_document" }, async () => {
    const summary = await summarize(document);
    const keywords = await extractKeywords(document);
    return { summary, keywords };
  });
}

// Run it
const result = await processDocument("Your document text here...");

​
Class Decorators

For TypeScript projects with experimentalDecorators enabled: 
import { workflow, task } from "@anyway-sh/node-server-sdk";

class DocumentProcessor {
  @workflow({ name: "process" })
  async process(document: string) {
    const summary = await this.summarize(document);
    const keywords = await this.extractKeywords(document);
    return { summary, keywords };
  }

  @task({ name: "summarize" })
  async summarize(text: string) {
    // LLM call here
  }

  @task({ name: "extract_keywords" })
  async extractKeywords(text: string) {
    // LLM call here
  }
}

​
Agent Tracing

Use withAgent to trace agent operations. Agent names automatically propagate to all child spans: 
import { withWorkflow, withAgent, withTask } from "@anyway-sh/node-server-sdk";

async function runAgent(query: string) {
  return withWorkflow({ name: "support_agent" }, async () => {
    return withAgent({ name: "classifier" }, async () => {
      const category = await withTask({ name: "classify" }, async () => {
        // LLM classification call
        return "billing";
      });

      const response = await withTask({ name: "draft_response" }, async () => {
        // LLM response generation
        return "Here is your response...";
      });

      return { category, response };
    });
  });
}

​
Association Properties

Attach metadata to traces for filtering and grouping in the dashboard: 
await withWorkflow(
  {
    name: "chat",
    associationProperties: {
      userId: "user-123",
      sessionId: "session-abc",
      environment: "production",
    },
  },
  async () => {
    // All spans in this workflow carry these properties
  },
);

​
Conversation Tracking

Track multi-turn conversations: 
import { withConversation } from "@anyway-sh/node-server-sdk";

async function handleMessage(conversationId: string, message: string) {
  return withConversation(conversationId, async () => {
    // All spans are tagged with the conversation ID
    const response = await client.chat.completions.create({
      model: "gpt-4o",
      messages: [{ role: "user", content: message }],
    });
    return response.choices[0].message.content;
  });
}

​
View in Dashboard

After running your code, view traces in the Anyway Dashboard:
  1. Navigate to Traces
  2. Find your workflow trace
  3. Expand to see nested task spans
  4. View timing, inputs, and outputs for each operation

​
Next Steps

Configuration

Configure endpoints and authentication

Customer & Order Attribution

Link traces to customers and orders

Cost Tracking

Monitor your AI spend