feat: Add SageMaker HyperPod EKS deployment target and MCP server int…#106
Merged
dferguson992 merged 1 commit intoawslabs:mainfrom Mar 17, 2026
Merged
Conversation
github-actions
bot
added
documentation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Issue #, if available: #82, #89
Description of changes:
Summary
Adds support for deploying ML containers to SageMaker HyperPod EKS clusters as an alternative to managed inference endpoints, and introduces MCP (Model Context Protocol) server integration for intelligent instance and region recommendations.
What changed
HyperPod EKS deployment target
New deploymentTarget prompt: users choose between managed-inference (existing) and hyperpod-eks
Kubernetes manifest templates (deployment.yaml, service.yaml, configmap.yaml, pvc.yaml) for HyperPod EKS deployments with GPU resource requests, tolerations, health probes on port 8080, and optional FSx volume mounts
Unified do/ scripts (deploy, clean, logs, test, config) now branch on deployment target — single script per action, no target-specific script files
do/deploy handles kubectl apply for HyperPod and aws sagemaker create-endpoint for managed inference
do/test supports kubectl port-forward testing for HyperPod and sagemaker-runtime invoke-endpoint for managed inference
do/export template added for configuration export
prompt-runner.js and prompts.js extended with HyperPod-specific prompts (cluster name, namespace, replicas, FSx volume handle)
template-manager.js handles conditional HyperPod directory inclusion via ignore patterns
config-manager.js extended with HyperPod parameter support, source enforcement, and deployment target awareness
MCP server integration
mcp-client.js and mcp-command-handler.js for managing MCP server configurations
Three MCP servers under servers/: instance-recommender, region-picker, and hyperpod-cluster-picker
Shared Bedrock client library in servers/lib/
MCP configuration documentation (
mcp-configuration.md
)
Test coverage
13 new property-based test files covering deployment target branching, K8s manifest correctness, script generation, prompt ordering, backward compatibility, and HyperPod cluster picker
Existing tests updated for the new deploymentTarget parameter
486 tests passing, 0 failing