Bug: TypeError: (0 , MGt.setCredentialFeature) is not a function

Expected Behaviour

https://awslabs.github.io/multi-agent-orchestrator/agents/built-in/lambda-agent/

Following instruction from this link should work.

Current Behaviour

import { MultiAgentOrchestrator } from "multi-agent-orchestrator"; const orchestrator = new MultiAgentOrchestrator(); orchestrator.addAgent(agent); const response = await orchestrator.routeRequest( "I need help with my order", "user123", "session456" );

processing request: TypeError: (0 , MGt.setCredentialFeature) is not a function at /var/task/index.js:12:34655 at e (/var/task/index.js:35:243) at /var/task/index.js:1:38819 at coalesceProvider (/var/task/index.js:1:38997) at /var/task/index.js:1:39199 at /var/task/node_modules/@smithy/core/dist-cjs/index.js:82:23 at async /var/task/index.js:1:7657 at async Oee.processRequest (/var/task/index.js:158:4342) at async Oee.classify (/var/task/index.js:157:76) at async o1e.routeRequest (/var/task/index.js:249:5168)

Code snippet

import {
  MultiAgentOrchestrator,
  LambdaAgent,
} from "multi-agent-orchestrator";
this.multiAgentOrchestrator = new MultiAgentOrchestrator();
   const lambdaAgents = [ {
        name: "system-generator",
        description: `
        An AI System Generator Agent to analyze business requirements and generate high-level system blueprints, breaking down complex processes into interconnected workflows.
        `,
        functionName: "system-generator",
        functionRegion: "ap-south-1",
      },
      {
        name: "workflow-generator",
        description: `
        An AI workflow Generator Agent to generate comprehensive, well-structured workflows based on given business cases or requirements.

        functionName: "ai-workflow-generator",
        functionRegion: "ap-south-1",
      },
    ];

    lambdaAgents.forEach((agentConfig) => {
      const agent = new LambdaAgent(agentConfig);
      this.multiAgentOrchestrator.addAgent(agent);
    });

  async processRequest(request: OrchestratorRequest): Promise<any> {
    const { input, userId, sessionId } = request;
    console.log('processRequest input, userId, sessionId:', input, userId, sessionId);

    try {
      const agentResponse = await this.multiAgentOrchestrator.routeRequest(
        "HR team manage their organization’s recruitment process",
        "user123",
        "session456"
      );
      console.log("agentResponse in AI Orchestrator:", agentResponse);

      return {
        result: "Request processed successfully",
        agentResponse,
      };
    } catch (error) {
      console.error("Error in AI Orchestrator:", error);
      throw error;
    }
  }

CDK code:
const aiOrchestrator = (
  scope: Construct,
  context: CDKContext,
  layers: lambda.ILayerVersion[]
) => {
  const handlerName = context.lambda.aiOrchestrator.name;
  const entryPath = join(
    __dirname,
    `/../../src/lambda/${handlerName}/src/index.js`
  );
  return createNodeJsFunction(
    scope,
    handlerName,
    context,
    entryPath,
    {
      layers: [...layers],
      memorySize: 512,
      environment: {
        ACCOUNT_NUMBER: context.accountNumber,
        REGION: context.region,
      },
      timeout: Duration.minutes(3),
    },
    {
      externalModulesArray: [
        // "@aws-lambda-powertools/logger",
        // "@aws-lambda-powertools/parameters",
      ],
      nodeModulesArray: [
        "@aws-sdk/client-cognito-identity-provider",
        "@aws-sdk/client-sso",
        "@aws-sdk/client-sts",
        "@smithy/core",
        "@aws-sdk/core",
      ],
    }
  );
};

export const createNodeJsFunction = (
  scope: Construct,
  handlerName: string,
  context: CDKContext,
  entry: string,
  props: NodejsFunctionProps | undefined,
  dependencies: Dependencies
): NodejsFunction => {
  console.log(
    "lambda path: ",
    __dirname,
    join(__dirname, `/../../src/lambda/${handlerName}/index.ts`)
  );
  return new NodejsFunction(scope, handlerName, {
    entry: join(entry),
    handler: "handler",
    // functionName: `${handlerName}-${context.appName}-${context.environment}`,
    functionName: `${handlerName}-${context.appName}`,
    runtime: Runtime.NODEJS_16_X,
    timeout: Duration.minutes(3),
    bundling: {
      externalModules: [...dependencies.externalModulesArray],
      nodeModules: [...dependencies.nodeModulesArray], // Bundle AWS SDK v3 modules
      minify: true, // Optional: Minimizes code
      // target: "node18", // Set target as Node.js 18
      // externalModules: [...externalModulesArray]
    },
    ...props,
  });
};

Bedrock Permission:
  const bedrockPolicy = new iam.PolicyStatement({
    actions: ["bedrock:InvokeModel", "bedrock:InvokeModelWithResponseStream"], // Allow invocation of Bedrock models
    //arn:aws:bedrock:ap-south-1::foundation-model/anthropic.claude-3-haiku-20240307-v1:0
    resources: [`arn:aws:bedrock:${context.region}::foundation-model/*`], // Adjust the region, account ID, and model as needed
    // resources: [`arn:aws:bedrock:${context.region}:${context.accountNumber}:model/*`], // Adjust the region, account ID, and model as needed
  });

Package.json:
{
  "name": "ai-orchestration-making",
  "version": "1.0.0",
  "description": "AI orchestration Lambda",
  "main": "index.js",
  "scripts": {
    "test": "jest"
  },
  "dependencies": {
    "@aws-sdk/client-cognito-identity-provider": "^3.675.0",
    "@aws-sdk/client-sso": "^3.675.0",
    "@aws-sdk/client-sts": "^3.675.0",
    "@aws-sdk/core": "^3.666.0",
    "@smithy/core": "^2.4.8",
    "aws-lambda": "^1.0.7",
    "multi-agent-orchestrator": "0.0.17"
  },
  "devDependencies": {
    "@types/aws-lambda": "8.10.145",
    "@types/jest": "^29.5.12",
    "jest": "^29.7.0",
    "ts-jest": "^29.1.2"
  }
}

Possible Solution

No response

Steps to Reproduce

try to run the code in a lambda runtime. Tried in both node16 and 18. Both have same issues.

awslabs / multi-agent-orchestrator