ECS MCP Server Sample
The following sample demonstrates how to deploy an Agentic AI architecture using AWS Fargate for Amazon ECS. It uses AWS CDK to deploy the architecture components.
接入你的智能体
從發布者安裝。不是 PlugMCP 端點。
遠端 URL 還是本機指令? 兩種安裝方式的差異
來自發布者
The following sample demonstrates how to deploy an Agentic AI architecture using AWS Fargate for Amazon ECS. It uses AWS CDK to deploy the architecture components.
This sample demonstrates how to deploy an Agentic AI architecture using AWS Fargate for Amazon ECS with AWS CDK. The sample features an AI Service that connects to multiple Model Context Protocol (MCP) servers to perform various actions

The architecture consists of the following components:
Service-to-service communication between AWS Fargate services is done using Service Connect. The AI Service leverages custom DNS names to connect to the MCP Servers available in this sample.
The sample also generates an AWS Secret to be used as the x-api-key header when calling the Application Load Balancer URLs.
To deploy this sample in your account follow the instructions below:
# install dependencies
$ npm install
# login to the public ECR registry to allow download base images
$ aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws
# bootstrap your AWS environment (in case you haven't done it yet)
$ npm run cdk bootstrap
# deploy the CDK Stack into your account
$ npm run cdk deploy
Once the deployment is finalized in your account, you will have the following output from AWS CDK:
✅ EcsMcpServerSampleStack
✨ Deployment time: 362.32s
Outputs:
EcsMcpServerSampleStack.ALBLoadBalancerDNS6ED38D87 = [application-load-balancer-url]
EcsMcpServerSampleStack.AiServiceSecretARN2FB050DB = [secret-arn]
Before invoking the Application Load Balancer URL, you'll need to retrieve the secret value with the following command:
$ aws secretsmanager get-secret-value --secret-id [secret-arn]
# or use the following to directly get the value
$ aws secretsmanager get-secret-value --secret-id [secret-arn] --output json | jq -r '.SecretString'
Now that you have all the information, you can call the Application Load Balancer URL to test the agent:
curl --location 'http://[ALB-URL]/chat' \
--header 'x-api-key: [secret-value]' \
--header 'Content-Type: application/json' \
--data '{
"prompt": "what'\''s the command to list AWS Buckets"
}'
{
"answer": "The command to list AWS S3 buckets is:\n\n```bash\naws s3 ls\n```\n\nThis command will display all S3 buckets in your AWS account, showing:\n- Creation date and time\n- Bucket name\n\nAs you can see from the output, it lists all the buckets with their creation timestamps and names. The command works across all regions and shows buckets from your entire AWS account.\n"
}
or execute the AWS CLI command:
curl --location 'http://[ALB-URL]/chat' \
--header 'x-api-key: [secret-value]' \
--header 'Content-Type: application/json' \
--data '{
"prompt": "list all my AWS Buckets"
}'
{
"answer": "Here are all your S3 buckets: [omissis]"
}
or compute some math using the custom MCP Server:
curl --location 'http://[ALB-URL]/chat' \
--header 'x-api-key: [secret-value]' \
--header 'Content-Type: application/json' \
--data '{
"prompt": "what'\''s 2+2"
}'
{
"answer": "2 + 2 = **4**\n"
}
See CONTRIBUTING for more information.
This library is licensed under the MIT-0 License. See the LICENSE file.
This sample demonstrates how to deploy an Agentic AI architecture using AWS Fargate for Amazon ECS with AWS CDK. The sample features an AI Service that connects to multiple Model Context Protocol (MCP
總覽
This sample demonstrates how to deploy an Agentic AI architecture using AWS Fargate for Amazon ECS with AWS CDK. The sample features an AI Service that connects to multiple Model Context Protocol (MCP) servers to perform various actions
工具
尚無工具清單。遠端伺服器會在首次健康檢查後取得工具清單。
常見問題
在本頁複製適用於 Cursor、Claude Desktop、Claude Code 或 VS Code 的 stdio 設定。它使用發布者的安裝方式,而不是 PlugMCP 代理伺服器。
這筆記錄未列出必要的環境變數。
Sample Ecs 透過 stdio 在本機執行。
上次檢查時間:2026-10-28。 This sample demonstrates how to deploy an Agentic AI architecture using AWS Fargate for Amazon ECS with AWS CDK. The sample features an AI Service that connects to multiple Model Context Protocol (MCP) servers to perform various actions