---
title: "Bedrock AgentCore MCP: review, setup & limitations · Undominated.ai"
canonical: https://undominated.ai/mcp-servers/awslabs-bedrock-agentcore/
description: "AWS Labs MCP server whose README says it exposes AgentCore runtime, memory, identity, gateway, policy, browser, and code-interpreter operations through boto3."
---

# Bedrock AgentCore MCP: review, setup & limitations · Undominated.ai

> AWS Labs MCP server whose README says it exposes AgentCore runtime, memory, identity, gateway, policy, browser, and code-interpreter operations through boto3.

[← Explore all mcp servers](/mcp-servers/)

CLOUD OPERATIONS / AWS Labs

# Bedrock AgentCore MCP

AWS Labs MCP server whose README says it exposes AgentCore runtime, memory, identity, gateway, policy, browser, and code-interpreter operations through boto3.

 [See setup guidance ↓](#setup)[Original source ↗](https://github.com/awslabs/mcp/tree/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server)

SOURCE REVIEW

 Reviewed 2026-10-14
 Evidence 2 linked sources
 Publisher AWS Labs
 Licence [Apache-2.0 ↗](https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/LICENSE)
 Revision 8840003820a7
 [Read what was—and wasn’t—checked ↓](#review)

“Lets a coding agent manage AgentCore resources, and the README's own table sums the surface as 122 tools across those primitives plus two documentation tools.”

 [AWS Labs · upstream description ↗](https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/README.md) Our analysis follows below.

01 / THE REASONING

## Why this made the selection.

 - The catalogue has no Bedrock AgentCore server.
- The README names browser and code-interpreter groups, which existing AWS servers do not cover.

### A good fit for

 - Managing AgentCore runtimes and gateways from a coding agent
- Searching the AgentCore docs tools named in the overview

### Weigh up before choosing

 - This review did not execute it.
- This review did not start the server or list its tools.
- The count of 122 tools is the README's table, not a recount of source symbols.
- The README does not use the words stdio or http in the portion reviewed, so no transport is recorded.

02 / THE REVIEW RECORD

## What we actually inspected.

Source review has boundaries. A clear record is more useful than a “safe” badge.

### Material inspected

 - src/amazon-bedrock-agentcore-mcp-server/README.md
- src/amazon-bedrock-agentcore-mcp-server/LICENSE

### Our findings

 - The overview says agents can create runtimes, store memories, configure identity, deploy gateways, and manage policies through boto3.
- The README table lists Runtime 14, Memory 21, Identity 21, Gateway 15, Policy 15, Browser 25, Code Interpreter 10, and Documentation 2, and states 122 tools. This review did not add those rows independently.
- Browser and code-interpreter tools are part of that table, so the surface is broader than resource reads. The directory LICENSE is Apache-2.0.

### Not established by this review

 - The server was not started.
- No AgentCore API was called.
- tools/list was not called.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

## Connect a server deliberately.

[Upstream setup instructions ↗](https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/README.md)
 - Install uv and Python 3.10.
- The client examples pass awslabs.amazon-bedrock-agentcore-mcp-server@latest to uvx inside JSON.
- No single shell line is recorded as the install command.

### Before you start

 - Python 3.10 and uv
- AWS credentials for AgentCore APIs

### Compatibility

Claude Code · Kiro · Cursor · VS Code · Codex CLI

### Implementation

python

### Broad AgentCore control

 - Create and manage runtimes, memory, identity, gateways, and policies
- Drive the cloud browser tools described in the table
- Execute code and move files in the code-interpreter tools described in the table

### Cost model

Apache-2.0; AWS credentials are required and were not used

04 / FOLLOW THE EVIDENCE

## The source trail.

Our notes are separate from the original resource. Check upstream before adopting a new version.

 - [AgentCore README ↗](https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/README.md) Checked 2026-10-14 https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/README.md Supports: primitive table and 122-tool claim
- [Directory LICENSE ↗](https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/LICENSE) Checked 2026-10-14 https://github.com/awslabs/mcp/blob/8840003820a78dd018561d55c7ab12fcdd708464/src/amazon-bedrock-agentcore-mcp-server/LICENSE Supports: Apache-2.0

KEEP COMPARING

## Other approaches to consider.

Related by category or shared topics. These are alternatives to inspect, not a measured quality order.

 [### AWS Billing and Cost Management MCP Server ↗ Reads realised AWS billing and cost-management data, which is a different number from the advertised unit price.](/mcp-servers/awslabs-aws-billing-cost-management/)[### AWS CloudTrail MCP Server ↗ Queries CloudTrail events for an account you administer.](/mcp-servers/awslabs-cloudtrail/)[### AWS CloudWatch Application Signals MCP Server ↗ Reads CloudWatch Application Signals for an account you administer. It is not the general CloudWatch server.](/mcp-servers/awslabs-cloudwatch-applicationsignals/)

 [AI Tools ↗](/tools/)[Skills ↗](/skills/)[Agents ↗](/agents/)[MCP Servers ↗](/mcp-servers/)[Workflows ↗](/workflows/)

## Continue your investigation

 - [Inspect resource evidence](/tools/)
- [Choose a host](/agents/)
- [Inspect reusable workflows](/skills/)
- [Use Undominated data](/api/)
