Merge pull request #2881 from burdeazy/burdeazy-feature-s3-lambda-agentcore

New pattern - s3-lambda-agentcore

Author: julianwood

s3-lambda-agentcore/.gitignore

@@ -0,0 +1,2 @@
+metadata
+invoke_agent.zip

s3-lambda-agentcore/Readme.md

@@ -0,0 +1,82 @@
+# Amazon S3 to AWS Lambda to Amazon Bedrock AgentCore
+This pattern creates an AWS Lambda function to invoke an agent in AgentCore Runtime when an object is uploaded to the Amazon S3 bucket.
+
+This Terraform template creates 2 S3 buckets (input and output), an AWS Lambda Function, and an agent in AgentCore Runtime.
+
+Learn more about this pattern at Serverless Land Patterns: https://serverlessland.com/patterns/s3-lambda-agentcore
+
+Important: this application uses various AWS services and there are costs associated with these services after the Free Tier usage - please see the AWS Pricing page for details. You are responsible for any AWS costs incurred. No warranty is implied in this example.
+
+## Requirements
+
+* [Create an AWS account](https://portal.aws.amazon.com/gp/aws/developer/registration/index.html) if you do not already have one and log in. The IAM user that you use must have sufficient permissions to make necessary AWS service calls and manage AWS resources.
+* [AWS CLI](https://docs.aws.amazon.com/cli/latest/userguide/install-cliv2.html) installed and configured
+* [Git Installed](https://git-scm.com/book/en/v2/Getting-Started-Installing-Git)
+* [Terraform](https://developer.hashicorp.com/terraform/tutorials/aws-get-started/install-cli) (Terraform) installed
+* [Docker](https://docs.docker.com/get-docker/) installed and running (required for building the agent container image)
+
+## Deployment Instructions
+
+1. Create a new directory, navigate to that directory in a terminal and clone the GitHub repository:
+
+`git clone https://github.com/aws-samples/serverless-patterns`
+
+2. Change directory to the pattern directory:
+
+`cd serverless-patterns/s3-lambda-agentcore`
+
+3. From the command line, initialize terraform to download and install the providers defined in the configuration:
+
+`terraform init`
+
+4. From the command line, apply the configuration in the deploy.tf file:
+
+`terraform apply`
+
+1. When prompted, enter `yes` to confirm the deployment
+
+2. Note the outputs from the deployment process, these contain the resource names and/or ARNs which are used for testing.
+
+## How it works 
+
+S3 will invoke the Lambda function when an object is created or updated. It will pass metadata about the new object in the event argument of the Lambda invocation.
+
+The lambda function will invoke the agent and pass a uri for the s3 file. 
+
+The agent will categorize the file as architecture, runbook, or other and identify some metadata. Then it will send the results back to the Lambda function as JSON. 
+
+The Lambda function will write the metadata to the S3 output bucket.
+
+## Testing
+
+Ensure you're in the correct directory (`cd serverless-patterns/s3-lambda-agentcore`). Then run the following script to test with files in the `./test-files` folder. 
+
+```bash
+# upload test files to the input bucket
+aws s3 cp ./test-files/ s3://$(terraform output -raw s3_input_bucket)/ --recursive
+# wait for the agent to process the files
+sleep 10
+# download the metadata from the output bucket
+aws s3 cp s3://$(terraform output -raw s3_output_bucket)/ ./metadata/  --recursive
+```
+You can view the metadata in `./metadata`
+
+## Cleanup
+
+1. Ensure you're in the correct directory (`cd serverless-patterns/s3-lambda-agentcore`)
+
+2. Delete all created resources: 
+
+`terraform destroy`
+
+3. When prompted, enter `yes` to confirm the destruction
+
+4. Confirm all created resources has been deleted: 
+
+`terraform show`
+
+---
+
+Copyright 2026 Amazon.com, Inc. or its affiliates. All Rights Reserved.
+
+SPDX-License-Identifier: MIT-0

s3-lambda-agentcore/agent/Dockerfile

@@ -0,0 +1,8 @@
+FROM python:3
+
+COPY requirements.txt ./
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+CMD [ "opentelemetry-instrument", "python", "main.py" ]

s3-lambda-agentcore/agent/main.py

@@ -0,0 +1,85 @@
+from strands import Agent
+from strands_tools import use_aws, current_time
+from strands.models import BedrockModel
+from bedrock_agentcore.runtime import BedrockAgentCoreApp
+from pydantic import BaseModel, Field
+from typing import List, Literal
+from datetime import datetime, timezone
+
+app = BedrockAgentCoreApp()
+
+# Define structured output schema
+class FileMetadata(BaseModel):
+    filename: str = Field(description="The name of the file")
+    system: str = Field(description="The system or service the file relates to")
+    keywords: List[str] = Field(description="List of relevant keywords or subjects")
+
+class FileClassification(BaseModel):
+    category: Literal["architecture", "operations", "other"] = Field(description="The category of the file")
+    metadata: FileMetadata = Field(description="Metadata about the file")
+    reasoning: str = Field(description="The reasoning behind the categorization")
+    time: str = Field(description="The UTC timestamp of the categorization")
+
+model_id = "us.amazon.nova-pro-v1:0"
+model = BedrockModel(
+    model_id=model_id,
+)
+
+agent = Agent(
+    model=model,
+    tools=[use_aws, current_time],
+    system_prompt="""
+You are an IT documentation classifier. Your task is to categorize documentation files into one of three categories and extract relevant metadata.
+
+CATEGORIES:
+
+1. **architecture** - System design and technical architecture documentation including:
+   - System architecture diagrams and design documents
+   - Reference architectures
+   - API specifications and interface definitions
+   - Data models, database schemas, and ER diagrams
+   - Technology stack decisions and architecture decision records (ADRs)
+   - Component interaction diagrams and sequence diagrams
+   - Infrastructure architecture and network topology
+   - Security architecture and authentication flows
+
+2. **operations** - Operational procedures and runbooks including:
+   - Deployment procedures and release processes
+   - Troubleshooting guides and incident response playbooks
+   - Monitoring and alerting setup documentation
+   - Backup and recovery procedures
+   - Configuration management and environment setup
+   - Maintenance schedules and operational checklists
+   - On-call procedures and escalation paths
+
+3. **other** - All other documentation including:
+   - Meeting notes and minutes
+   - Project plans and timelines
+   - Training materials and user guides
+   - General reference documents
+   - Administrative documentation
+
+TASK:
+
+For each file, analyze its content and provide:
+- **category**: One of "architecture", "operations", or "other"
+- **metadata**:
+  - **filename**: The name of the file
+  - **system**: The primary system, service, or component the document relates to
+  - **keywords**: A list of relevant technical keywords or topics covered
+
+Base your categorization on the document's primary purpose and content. If a document covers multiple areas, choose the category that best represents its main focus.
+"""
+)
+
+@app.entrypoint
+def strands_agent_bedrock(payload):
+    """
+    Invoke the agent with a payload and return structured output
+    """
+    user_input = payload.get("prompt")
+    response = agent(user_input, structured_output_model=FileClassification)
+    return response.structured_output.model_dump()
+
+if __name__ == "__main__":
+    app.run()

s3-lambda-agentcore/agent/requirements.txt

@@ -0,0 +1,7 @@
+strands-agents
+strands-agents-tools
+uv
+boto3
+bedrock-agentcore<=0.1.5
+bedrock-agentcore-starter-toolkit==0.1.14
+aws-opentelemetry-distro>=0.10.0

s3-lambda-agentcore/bin/build.sh

@@ -0,0 +1,41 @@
+#!/bin/bash
+
+# Fail fast
+set -e
+
+# This is the order of arguments
+ECR_BASE_ARN=${1}
+BUILD_FOLDER=${2}
+IMAGE_NAME=${3}
+IMAGE_URI=${4}
+TARGET_AWS_REGION=${5}
+MYTAG=$(date +%Y%m%d%H%M%S)
+
+# Check that aws is installed
+which aws >/dev/null || {
+    echo 'ERROR: aws-cli is not installed'
+    exit 1
+}
+
+# Check that docker is installed and running
+which docker >/dev/null && docker ps >/dev/null || {
+    echo 'ERROR: docker is not running'
+    exit 1
+}
+
+# Connect into aws
+aws ecr get-login-password --region ${TARGET_AWS_REGION} | docker login --username AWS --password-stdin ${ECR_BASE_ARN} || {
+    echo 'ERROR: aws ecr login failed'
+    exit 1
+}
+
+# Build image
+docker build --no-cache -t ${IMAGE_NAME} ${BUILD_FOLDER} --platform linux/arm64
+
+# Docker Tag and Push
+docker tag ${IMAGE_NAME}:latest ${IMAGE_URI}:latest
+docker tag ${IMAGE_URI}:latest ${IMAGE_URI}:${MYTAG}
+docker push ${IMAGE_URI}:latest
+docker push ${IMAGE_URI}:${MYTAG}
+
+echo "Tags Used for ${IMAGE_NAME} Image are ${MYTAG}"