Model Context Protocol: construct(s) to facilitate the creation of MCP servers on AWS
Describe the feature
Anthropic released the https://www.anthropic.com/news/model-context-protocol . This is an open standard that enables developers to build secure, two-way connections between their data sources and AI-powered tools.
The idea of the new construct(s) would be to facilitate the creation of new MCP servers on AWS.
More info/resources:
- https://github.com/awslabs/mcp
- https://github.com/modelcontextprotocol/servers
- https://modelcontextprotocol.io/development/roadmap
- https://github.com/aws-samples/sample-mcp-server-s3
- https://github.com/mikegc-aws/amazon-bedrock-mcp
Use Case
Facilitate how to give Large Language Models (LLMs) secure, controlled access to tools and data sources.
Proposed Solution
No response
Other Information
No response
Acknowledgements
- [ ] I may be able to implement this feature request
- [ ] This feature might incur a breaking change
Tagging @laithalsaadoon @spugachev to get your thoughts on this
Proposed Solution
This construct should deploy an MCP server on AWS ECS using Fargate. It should include an Application Load Balancer and support Docker-based MCP servers from public Docker-compatible registries. The construct should be able use an existing VPC or create a new one if VPC is not specified.
Features
- Deploy containerized MCP servers on AWS ECS.
- Support for public Docker-compatible registries (DockerHub, Amazon ECR Public, GitHub Container Registry, etc.).
- Automatic scaling and load balancing.
- Flexible configuration options (e.g., task count, CPU/memory allocation).
- Option to use an existing VPC or create a new one automatically.
This issue is now marked as stale because it hasn't seen activity for a while. Add a comment or it will be closed soon. If you wish to exclude this issue from being marked as stale, add the "backlog" label.
Closing this issue as it hasn't seen activity for a while. Please add a comment @mentioning a maintainer to reopen. If you wish to exclude this issue from being marked as stale, add the "backlog" label.