Amazon Bedrock Agent Runtime JavaScript

Instrument and observe AWS Bedrock Agent Runtime calls in JavaScript/Node.js

This module provides automatic instrumentation for the AWS SDK for JavaScript Bedrock Agent Runtime Client, which may be used in conjunction with @arizeai/phoenix-otel.

Install

npm install @arizeai/openinference-instrumentation-bedrock-agent-runtime @aws-sdk/client-bedrock-agent-runtime @arizeai/phoenix-otel

Setup

To instrument your application, use the register function from @arizeai/phoenix-otel and manually instrument the Bedrock Agent Runtime SDK.

Create the instrumentation.ts file:

import { register } from "@arizeai/phoenix-otel";
import { BedrockAgentRuntimeClient } from "@aws-sdk/client-bedrock-agent-runtime";
import { BedrockAgentRuntimeInstrumentation } from "@arizeai/openinference-instrumentation-bedrock-agent-runtime";

// Initialize Phoenix tracing
const tracerProvider = register({
  projectName: "bedrock-agent-app",
  // If Phoenix is running elsewhere:
  // url: "https://your-phoenix.example.com",
  // apiKey: process.env.PHOENIX_API_KEY,
  // If using self-hosted Phoenix:
  // url: "http://localhost:6006",
});

// Set up Bedrock Agent Runtime instrumentation
const instrumentation = new BedrockAgentRuntimeInstrumentation();
instrumentation.manuallyInstrument(BedrockAgentRuntimeClient);

console.log("Bedrock Agent Runtime instrumentation registered");

Run Bedrock Agents

Import the instrumentation.ts file first, then use Bedrock Agent Runtime as usual.

import "./instrumentation.js";
import { BedrockAgentRuntimeClient, InvokeAgentCommand } from "@aws-sdk/client-bedrock-agent-runtime";

const client = new BedrockAgentRuntimeClient({ region: "us-east-1" });

async function main() {
  const command = new InvokeAgentCommand({
    agentId: "YOUR_AGENT_ID",
    agentAliasId: "YOUR_AGENT_ALIAS_ID",
    sessionId: `session-${Date.now()}`,
    inputText: "What's the weather like today?",
  });

  const response = await client.send(command);
  
  // Process streaming response
  if (response.completion) {
    for await (const event of response.completion) {
      if (event.chunk?.bytes) {
        const text = new TextDecoder().decode(event.chunk.bytes);
        console.log(text);
      }
    }
  }
}

main();

Observe

After setting up instrumentation and running your Bedrock Agent application, traces will appear in the Phoenix UI for visualization and analysis. This includes:

• Agent invocations
• Action group calls (as tools)
• Knowledge base lookups
• LLM calls within the agent

Resources

• NPM Package

• OpenInference package for AWS Bedrock Agent Runtime