bedrock-agentcore-agents

Source-cited best-practices playbook for building production AI agents on AWS that maps the customer's intent and use case to the right architecture. Covers Strands Agents, Amazon Bedrock (model choice, Converse API, Guardrails, Knowledge Bases/RAG), and Amazon Bedrock AgentCore (Runtime, Memory, Gateway, Identity, Policy/Cedar, built-in Browser and Code Interpreter tools), plus multi-agent patterns, governance, Terraform-first IaC, cost/scaling, and CloudWatch/OpenTelemetry observability. Use whenever the user wants to choose, design, build, deploy, secure, govern, monitor, or debug an AI agent on AWS, even if no service is named. Triggers: build an agent on Bedrock, Strands agent, AgentCore, deploy to AWS, RAG, multi-agent, model choice, agent memory, MCP gateway, guardrails, governance, Bedrock IAM, observability, cost, Terraform for Bedrock, or pasted boto3 bedrock-runtime, bedrock-agentcore, or strands code to productionize or debug. Every recommendation cites an official source.

License

MIT-0