Agent building & orchestration
AWS's set of composable services for running agents built with any framework, covering serverless execution with isolated sessions, persistent memory, a gateway that turns APIs into MCP tools, identity and credential management, and built-in observability. Documented governance is strongest on identity and authorization rather than approvals.
commercial · generally available · Research snapshot 2026-09-06
Visit the official product source ↗Agent building & orchestration · Agent identity & access · Runtime authorization & controls
Useful conversation with: Cloud architect, Platform engineering lead, AI engineering manager.
Show me an AgentCore Gateway exposing two Lambda tools where the agent's inbound and outbound authorization is verified per request through AgentCore Identity.
Support labels reflect the supplied research. Documentation and vendor claims are not independent product tests. “Not established” means the researcher did not find support; it does not prove a capability is absent.
Documented by provider
Agents built with frameworks including CrewAI, LangGraph, LlamaIndex and Strands Agents can be deployed with serverless execution, isolated sessions, persistent memory and built-in observability, without infrastructure management.
Limit: The page does not state GA versus preview status for individual AgentCore services.
Documented by provider
Existing APIs, databases and services can be turned into tools for MCP-compatible agents, including a gateway that wraps Lambda functions or OpenAPI specifications without rewriting code.
Limit: Tool-level authorization granularity is not described on the overview page.
Documented by provider
AgentCore Identity implements authentication and authorization controls that verify each request independently and require explicit verification for all access attempts regardless of source.
Limit: The overview page does not document RBAC models, audit logging or human approval flows for agent actions.
Documented by provider
Provides internal developers a governed path to build and deploy agents using approved tools, shared memory stores and centralized observability, authentication and compliance.
Limit: 'Approved tools' and 'compliance' are stated at a high level without describing the approval or attestation mechanism.
Not established in this research
A built-in human-in-the-loop approval gate for agent actions is not established by the pages reviewed.
Limit: Absence of evidence; other AgentCore documentation pages were not fetched.
Listing does not imply partnership, supplier status, a working DutyGraph integration, or a compliance certification.
Suggest a correction