◈ AWS AGENTCORE BLUEPRINTS
PART II · CORE SERVICES · CHAPTER 11 / 36
TOCCONSULTING.FR
AMAZON BEDROCK / SERVICE DETAIL S04

AGENTCORE GATEWAY

CHAPTER 11 / 36
IN ONE LINE turns your APIs into agent tools
REV 2026.07

Agents need tools (functions and APIs) to do real work. Wiring each one up by hand is slow. Gateway does it for you: point it at an API or an AWS Lambda function and it becomes a tool the agent can use, behind one secure address, with the login and permissions handled.

New to these words? Read chapter 02 "Key terms" (tool, MCP, Lambda, OAuth).
[ 1 ] WHAT GATEWAY IS

Gateway is a single front door between your agent and your tools. It takes things you already have, your APIs and AWS Lambda functions, and exposes them as tools the agent can call using MCP (the standard way agents talk to tools). It is the only fully managed AWS service that handles security on both sides: who may use the gateway (inbound), and how the gateway logs in to your services (outbound).

YOU PLUG IN
OpenAPI (a description of a REST API), Smithy, Lambda functions, or existing MCP servers.
THE AGENT GETS
One address with all your tools in a single list, plus search to find the right one.
[ 2 ] THE MAIN WAYS TO PLUG THINGS IN (A THIRD, INFERENCE TARGETS, ROUTES LLM TRAFFIC ACROSS MODEL PROVIDERS)
COMBINE MANY INTO ONE aggregation
Add several backends: Lambda, OpenAPI, Smithy, MCP servers.
Gateway merges them into one tool list the agent sees.
Includes search so the agent can find the right tool among many.
PASS STRAIGHT THROUGH proxy
For a web service such as an AgentCore Runtime agent.
Traffic goes directly to it, with no merging.
Each is reached on its own path.
[ 3 ] WHAT HAPPENS WHEN THE AGENT CALLS A TOOL
STEP 01

FIND

The agent reads the tool list, or searches it by meaning to pick the right tool.

==>
STEP 02

CHECK CALLER

Gateway checks who is calling (inbound auth) before letting the request through.

==>
STEP 03

TRANSLATE

It turns the tool call into a real API request or a Lambda run.

==>
STEP 04

LOG IN

It adds the right credentials to reach your service (outbound auth).

==>
STEP 05

RETURN

It hands the result back to the agent as a tool answer.

[ 4 ] SECURITY ON BOTH SIDES
INBOUND who may use the gateway
OAuth (a login token), or AWS SigV4 (AWS identity).
Check only: confirm the token, let the tool decide the rest.
None: open, for quick testing only.
OUTBOUND how the gateway reaches your service
For Lambda: an AWS role you grant it (Smithy: a role, or OAuth).
For OpenAPI and MCP servers: a stored API key or OAuth login.
Or AWS SigV4, or none for public services.
[ 5 ] WHY TEAMS USE IT
ONE ADDRESS

Many APIs and functions behind a single secure endpoint.

FIND THE TOOL

Search lets an agent use thousands of tools without a giant prompt.

READY-MADE

One-click links to Salesforce, Slack, Jira, Asana, Zendesk.

NO SERVERS

Fully managed, scales itself, with built-in logging.

Plain definitions: chapter 02 "Key terms". Hands-on commands and code: chapter 12 "Gateway, practical".
SRC: AWS AgentCore Developer Guide (gateway, core concepts, supported targets). DRAWN 2026.06, REV 2026.07.26
◄ PREVIOUS · CH 10
AgentCore Memory, Practical
NEXT · CH 12 ►
AgentCore Gateway, Practical
AWS AGENTCORE BLUEPRINTS · CONTENTS
Back to tocconsulting.fr Cover & Table of Contents
PART I · FOUNDATIONS
01AgentCore Overview, Plain LanguageCONCEPT02AgentCore, Key Terms in Plain LanguageCONCEPT03AgentCore, Common Mistakes ExplainedCONCEPT04Inside an AgentCore Agent, BlueprintCONCEPT05Harness, Practical 1, Strands PathPRACTICAL06Harness, Practical 2, Managed HarnessPRACTICAL
PART II · CORE SERVICES
07AgentCore Runtime, Plain LanguageCONCEPT08AgentCore Runtime, PracticalPRACTICAL09AgentCore Memory, Plain LanguageCONCEPT10AgentCore Memory, PracticalPRACTICAL11AgentCore Gateway, Plain LanguageCONCEPT12AgentCore Gateway, PracticalPRACTICAL13AgentCore Identity, Plain LanguageCONCEPT14AgentCore Identity, PracticalPRACTICAL
PART III · TOOLS AND OPERATIONS
15AgentCore Code Interpreter, Plain LanguageCONCEPT16AgentCore Code Interpreter, PracticalPRACTICAL17AgentCore Browser, Plain LanguageCONCEPT18AgentCore Browser, PracticalPRACTICAL19AgentCore Observability, Plain LanguageCONCEPT20AgentCore Observability, PracticalPRACTICAL21AgentCore Evaluations, Plain LanguageCONCEPT22AgentCore Evaluations, PracticalPRACTICAL
PART IV · GOVERNANCE AND TRANSACTIONS
23AgentCore Policy, Plain LanguageCONCEPT24AgentCore Policy, PracticalPRACTICAL25AgentCore Registry, Plain LanguageCONCEPT26AgentCore Registry, PracticalPRACTICAL27AgentCore Payments, Plain LanguageCONCEPT28AgentCore Payments, PracticalPRACTICAL
PART V · MULTI-AGENT AND ARCHITECTURE
29A2A and Multi-Agent, Plain LanguageCONCEPT30A2A and Multi-Agent, PracticalPRACTICAL31Architecture Patterns, Plain LanguageCONCEPT32Multi-Tenant Agents, Plain LanguageCONCEPT33Agent Security and Threat Model, Plain LanguageCONCEPT34Governance at Org Scale, Plain LanguageCONCEPT35Governance at Org Scale, PracticalPRACTICAL36Multi-Region and Distribution, Plain LanguageCONCEPT