feat: add AWS Lambda Managed Instances (LMI) skill to aws-serverless plugin

README.md

@@ -32,7 +32,7 @@ To maximize the benefits of plugin-assisted development while maintaining securi
 | --------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------- |
 | **amazon-location-service** | Add maps, geocoding, routing, places search, and geospatial features to applications with Amazon Location Service                                                                                       | Available                             |
 | **aws-amplify**             | Build full-stack apps with AWS Amplify Gen 2 using guided workflows for auth, data, storage, and functions                                                                                              | Available                             |
-| **aws-serverless**          | Build serverless applications with Lambda, API Gateway, EventBridge, Step Functions, and durable functions                                                                                              | Available                             |
+| **aws-serverless**          | Build serverless applications with Lambda, API Gateway, EventBridge, Step Functions, durable functions, and Lambda Managed Instances                                                                    | Available                             |
 | **databases-on-aws**        | Database guidance for the AWS database portfolio — schema design, queries, migrations, and multi-tenant patterns                                                                                        | Some Services Available (Aurora DSQL) |
 | **deploy-on-aws**           | Deploy applications to AWS with architecture recommendations, cost estimates, and IaC deployment                                                                                                        | Available                             |
 | **migration-to-aws**        | Migrate GCP infrastructure to AWS with resource discovery, architecture mapping, cost analysis, and execution planning                                                                                  | Available                             |
@@ -205,7 +205,7 @@ Build full-stack apps with AWS Amplify Gen 2 using TypeScript code-first develop
 
 ## aws-serverless
 
-Design, build, deploy, test, and debug serverless applications with AWS Lambda, API Gateway, EventBridge, Step Functions, and durable functions. Includes SAM and CDK deployment workflows, a SAM template validation hook, and the AWS Lambda durable functions skill for building resilient, long-running, multi-step applications.
+Design, build, deploy, test, and debug serverless applications with AWS Lambda, API Gateway, EventBridge, Step Functions, and durable functions. Includes SAM and CDK deployment workflows, a SAM template validation hook, the AWS Lambda durable functions skill for building resilient, long-running, multi-step applications, and the Lambda Managed Instances skill for evaluating, configuring, and migrating workloads to EC2-backed Lambda.
 
 ### Agent Skill Triggers
 
@@ -214,6 +214,7 @@ Design, build, deploy, test, and debug serverless applications with AWS Lambda,
 | **aws-lambda**                   | "Lambda function", "event source", "serverless application", "API Gateway", "EventBridge", "Step Functions", "serverless API", "event-driven architecture", "Lambda trigger" |
 | **aws-serverless-deployment**    | "use SAM", "SAM template", "SAM init", "SAM deploy", "CDK serverless", "CDK Lambda construct", "NodejsFunction", "PythonFunction", "serverless CI/CD pipeline"               |
 | **aws-lambda-durable-functions** | "lambda durable functions", "workflow orchestration", "state machines", "retry/checkpoint patterns", "long-running stateful Lambda", "saga pattern", "human-in-the-loop"     |
+| **aws-lambda-managed-instances** | "Lambda Managed Instances", "LMI", "capacity provider", "multi-concurrency Lambda", "EC2-backed Lambda", "cold start elimination", "Graviton Lambda", "Lambda cost optimization with Reserved Instances" |
 
 ### MCP Servers
 

plugins/aws-serverless/.claude-plugin/plugin.json

@@ -8,6 +8,8 @@
     "aws",
     "lambda",
     "durable functions",
+    "managed-instances",
+    "lmi",
     "serverless",
     "development",
     "sam",

plugins/aws-serverless/skills/aws-lambda-managed-instances/SKILL.md

@@ -0,0 +1,205 @@
+---
+name: aws-lambda-managed-instances
+description: >
+  Evaluate, configure, and migrate workloads to AWS Lambda Managed Instances (LMI).
+  Triggers on: Lambda Managed Instances, LMI, capacity provider, multi-concurrency Lambda,
+  dedicated instance Lambda, EC2-backed Lambda, cold start elimination, Graviton Lambda,
+  instance type for Lambda, Lambda cost optimization with Reserved Instances or Savings Plans.
+  Also trigger when users describe high-volume predictable workloads seeking cost savings,
+  or compare Lambda vs EC2 for steady-state traffic. For standard Lambda without LMI,
+  use the aws-lambda skill instead.
+argument-hint: "[describe your workload or what you need help with]"
+metadata:
+  tags: lambda, lmi, managed-instances, ec2, capacity-provider, multi-concurrency, cost-optimization
+---
+
+# AWS Lambda Managed Instances (LMI)
+
+Run Lambda functions on current-generation EC2 instances in your account while AWS manages provisioning, patching, scaling, routing, and load balancing. Combines Lambda's developer experience with EC2's pricing and hardware options.
+
+For standard Lambda development, see [aws-lambda skill](../aws-lambda/). For SAM/CDK deployment, see [aws-serverless-deployment skill](../aws-serverless-deployment/).
+
+## When to Load Reference Files
+
+- **Cost comparison**, **pricing analysis**, **Lambda vs LMI cost**, **Savings Plans**, or **Reserved Instances** -> see [references/cost-comparison.md](references/cost-comparison.md)
+- **Instance types**, **memory sizing**, **vCPU ratios**, **scaling tuning**, or **capacity provider config** -> see [references/configuration-guide.md](references/configuration-guide.md)
+- **Thread safety**, **code review checklist**, or **multi-concurrency readiness** -> see [references/thread-safety.md](references/thread-safety.md)
+- **Before/after code examples**, **runtime-specific migration** (Node.js, Python, Java, .NET), or **connection pooling** -> see [references/migration-patterns.md](references/migration-patterns.md)
+- **IAM roles**, **VPC setup**, **CLI commands**, **SAM template**, or **CDK example** -> see [references/infrastructure-setup.md](references/infrastructure-setup.md)
+- **Errors**, **throttling**, **debugging**, or **stuck deployments** -> see [references/troubleshooting.md](references/troubleshooting.md)
+
+## Quick Decision: Is LMI Right for This Workload?
+
+| Signal | LMI is a strong fit | Standard Lambda is better |
+|--------|---------------------|---------------------------|
+| Traffic | Steady, predictable, 50M+ req/mo | Bursty, unpredictable, long idle |
+| Cost | Duration-heavy spend at scale | Low or sporadic invocations |
+| Cold starts | Unacceptable (LMI has zero) | Tolerable or mitigated by SnapStart |
+| Compute | Latest CPUs, specific families, high network BW | Standard Lambda memory/CPU sufficient |
+| Compliance | Single-tenant required, VPC control | Multi-tenant Firecracker acceptable |
+| Scale-to-zero | Not needed (min 3 instances always run) | Required (pay nothing when idle) |
+| Code readiness | Thread-safe or feasible to refactor | Non-thread-safe, expensive to change |
+
+## Instructions
+
+### Step 1: Assess the Workload
+
+Gather these signals before recommending:
+
+1. **Traffic pattern**: Steady vs bursty? Requests per second?
+2. **Current costs**: Monthly Lambda spend? Existing Savings Plans?
+3. **Runtime**: Node.js, Java, .NET, or Python?
+4. **Memory/CPU**: How much memory? CPU-bound or I/O-bound?
+5. **Execution duration**: Average and P99?
+6. **Thread safety**: Mutable globals, shared `/tmp` paths, non-thread-safe libs?
+7. **VPC**: Already in a VPC? Private resource access needed?
+
+### Step 2: Build the Cost Comparison
+
+REQUIRED: Present a 4-column comparison before recommending LMI.
+
+| Scenario | When it wins |
+|----------|-------------|
+| Lambda on-demand | Low volume, bursty traffic |
+| Lambda + Savings Plan | Moderate steady volume (~17% duration discount) |
+| LMI on-demand | High volume, steady traffic |
+| LMI + 3yr Savings Plan | High volume + commitment (up to 72% EC2 discount) |
+
+Rule of thumb: LMI becomes cost-competitive at 50-100M+ req/month with steady traffic.
+
+See [references/cost-comparison.md](references/cost-comparison.md) for formulas, worked example, and comparison table template.
+
+### Step 3: Configure the Deployment
+
+**Instance families** (400+ types, .large and up): C-series (compute), M-series (general), R-series (memory). ARM (Graviton) for best price-performance.
+
+**Memory-to-vCPU ratios**: 2:1 (compute), 4:1 (general, default), 8:1 (memory). Min 2 GB, max 32 GB.
+
+**Multi-concurrency defaults/vCPU**: Node.js 64, Java 32, .NET 32, Python 16.
+
+**Scaling**: MinExecutionEnvironments (default 3), MaxVCpuCount (required), TargetResourceUtilization.
+
+See [references/configuration-guide.md](references/configuration-guide.md) for decision trees and detailed tuning.
+
+### Step 4: Migrate the Code
+
+Review code for thread safety. LMI runs multiple invocations concurrently per execution environment.
+
+**Common issues**: mutable globals, shared `/tmp` paths, non-thread-safe libs, per-invocation DB connections.
+
+See [references/thread-safety.md](references/thread-safety.md) for the review checklist and [references/migration-patterns.md](references/migration-patterns.md) for runtime-specific before/after code.
+
+### Step 5: Set Up Infrastructure
+
+Two IAM roles required (execution + operator). VPC with 3+ AZ subnets. Create capacity provider, attach function, publish version.
+
+See [references/infrastructure-setup.md](references/infrastructure-setup.md) for CLI commands, SAM, and CDK templates.
+
+### Step 6: Validate and Cut Over
+
+1. Test locally with LocalStack (supports LMI emulation)
+2. Monitor CloudWatch: CPU utilization, memory, concurrency, throttle rate
+3. Gradual traffic shift with weighted aliases (10% → 50% → 100%)
+4. Compare costs after 1-2 weeks of production data
+5. Decommission standard Lambda once stable
+
+## Best Practices
+
+### Configuration
+
+- Do: Start with 4:1 ratio and runtime default concurrency
+- Do: Use ARM (Graviton) unless x86 dependencies exist
+- Do: Let Lambda choose instance types unless specific hardware needed
+- Do: Set MaxVCpuCount to control cost ceiling
+- Don't: Set MinExecutionEnvironments below 3 (breaks AZ resiliency)
+- Don't: Over-restrict instance types (lowers availability)
+
+### Migration
+
+- Do: Review all code for thread safety before attaching to capacity provider
+- Do: Use weighted aliases for gradual traffic shift
+- Do: Include request IDs in all log statements
+- Do: Initialize DB pools and SDK clients outside the handler
+- Don't: Write to hardcoded `/tmp` paths without request-unique naming
+- Don't: Skip cost comparison — LMI is not always cheaper
+
+### Operations
+
+- Do: Set CloudWatch alarms on throttle rate > 1% and CPU > 80%
+- Do: Plan for 14-day instance rotation (automatic)
+- Don't: Manually terminate LMI EC2 instances (delete the capacity provider instead)
+- Don't: Forget to publish a version — unpublished functions cannot run on LMI
+
+## Limits Quick Reference
+
+| Resource | Limit |
+|----------|-------|
+| Memory | 2 GB min, 32 GB max |
+| Instances | 3 minimum (AZ resiliency) |
+| Instance lifespan | 14 days (auto-replaced) |
+| Concurrency/vCPU | 64 (Node.js), 32 (Java/.NET), 16 (Python) |
+| Runtimes | Node.js, Java, .NET, Python |
+| Instance families | C, M, R (.large and up) |
+| Scaling | Absorbs 50% spike; doubles within 5 min |
+
+## Troubleshooting Quick Reference
+
+| Issue | Cause | Fix |
+|-------|-------|-----|
+| 429 throttles | Traffic exceeds scaling speed | Increase MinExecutionEnvironments or lower TargetResourceUtilization |
+| Function stuck PENDING | Provisioning instances | Wait; check VPC/IAM config |
+| Architecture mismatch | Function ≠ capacity provider arch | Align both to same architecture |
+| Cannot terminate instances | Managed by capacity provider | Delete capacity provider instead |
+| Race conditions | Code not thread-safe | See [references/thread-safety.md](references/thread-safety.md) |
+
+See [references/troubleshooting.md](references/troubleshooting.md) for detailed resolution steps.
+
+## Configuration
+
+### AWS CLI Setup
+
+REQUIRED: AWS credentials configured on the host machine.
+
+**Verify access**: Run `aws sts get-caller-identity`
+
+### Regional Availability
+
+us-east-1, us-east-2, us-west-2, ap-northeast-1, eu-west-1
+
+## Language Selection
+
+Default: TypeScript
+
+Override: "use Python" → Python, "use JavaScript" → JavaScript. When not specified, ALWAYS use TypeScript.
+
+## IaC Framework Selection
+
+Default: CDK
+
+Override: "use SAM" → SAM YAML, "use CloudFormation" → CloudFormation YAML. When not specified, ALWAYS use CDK.
+
+## Error Scenarios
+
+### Serverless MCP Server Unavailable
+
+- Inform user: "AWS Serverless MCP not responding"
+- Ask: "Proceed without MCP support?"
+- DO NOT continue without user confirmation
+
+### Unsupported Runtime
+
+- State: "Lambda Managed Instances does not yet support [runtime]"
+- List supported runtimes
+- Suggest standard Lambda as alternative
+
+### Unsupported Region
+
+- State: "Lambda Managed Instances is not yet available in [region]"
+- List available regions
+
+## Resources
+
+- [Lambda Managed Instances Docs](https://docs.aws.amazon.com/lambda/latest/dg/lambda-managed-instances.html)
+- [Introducing LMI (AWS Blog)](https://aws.amazon.com/blogs/aws/introducing-aws-lambda-managed-instances-serverless-simplicity-with-ec2-flexibility/)
+- [Build High-Performance Apps with LMI](https://aws.amazon.com/blogs/compute/build-high-performance-apps-with-aws-lambda-managed-instances/)
+- [AWS Lambda Pricing](https://aws.amazon.com/lambda/pricing/)

plugins/aws-serverless/skills/aws-lambda-managed-instances/references/configuration-guide.md

@@ -0,0 +1,59 @@
+# LMI Configuration Guide
+
+## Instance Type Decision Tree
+
+- **CPU-intensive** (encoding, ML, compression) → C-series, 2:1 ratio, concurrency=1/vCPU
+- **Memory-intensive** (caching, large datasets) → R-series, 8:1 ratio
+- **Network-intensive** (streaming, data transfer) → Use AllowedInstanceTypes for n-suffix types, 4:1 ratio
+- **General/balanced** (web APIs, microservices) → M-series, 4:1 ratio, default concurrency
+
+Architecture: ARM (Graviton, g-suffix) for price-performance. x86 (i=Intel, a=AMD) when dependencies require it.
+
+## Memory-to-vCPU Ratios
+
+| Ratio | Profile | When to use | Memory examples |
+|-------|---------|-------------|-----------------|
+| 2:1 | Compute | CPU-bound work | 2GB/1vCPU, 4GB/2vCPU |
+| 4:1 | General | Most workloads (default) | 4GB/1vCPU, 8GB/2vCPU |
+| 8:1 | Memory | Caching, data, Python apps | 8GB/1vCPU, 16GB/2vCPU |
+
+Min: 2 GB / 1 vCPU. Max: 32 GB. Memory must align with ratio multiples.
+
+## Memory Sizing from Existing Lambda
+
+| Current Lambda | LMI memory | Ratio | Rationale |
+|---------------|------------|-------|-----------|
+| 128-512 MB | 2048 MB | 4:1 | LMI minimum; multi-concurrency shares memory |
+| 512 MB-1 GB | 2048 MB | 4:1 | Room for concurrent requests |
+| 1-2 GB | 4096 MB | 4:1 | Standard upgrade path |
+| 2-4 GB | 4096-8192 MB | 4:1 or 8:1 | Depends on memory vs CPU bottleneck |
+| 4-10 GB | 8192-16384 MB | 8:1 | Likely memory-heavy workload |
+
+## Concurrency Tuning
+
+| Runtime | Default/vCPU | I/O-bound | CPU-bound |
+|---------|-------------|-----------|-----------|
+| Node.js | 64 | Keep or increase | 1 per vCPU |
+| Java | 32 | Keep | 1 per vCPU |
+| .NET | 32 | Keep | 1 per vCPU |
+| Python | 16 | Keep | 1 per vCPU |
+
+Total capacity = MinExecutionEnvironments × PerExecutionEnvironmentMaxConcurrency
+
+## Capacity Provider Scaling Controls
+
+| Control | Default | Guidance |
+|---------|---------|----------|
+| MinExecutionEnvironments | 3 | Increase for baseline capacity; never below 3 |
+| MaxExecutionEnvironments | — | Set based on cost budget |
+| MaxVCpuCount | Required | Start at 30, adjust by load |
+| TargetResourceUtilization | ~50% headroom | Raise for cost savings (less burst tolerance) |
+| AllowedInstanceTypes | All | Restrict only for specific hardware needs |
+| ExcludedInstanceTypes | None | Exclude expensive types in dev/test |
+
+## Monitoring Thresholds
+
+- **CPU > 80%**: reduce concurrency or add vCPUs
+- **CPU < 20%**: increase concurrency for better utilization
+- **Throttle rate (429s) > 1%**: increase MinExecutionEnvironments or reduce utilization target
+- **Memory > 90%**: increase memory or reduce concurrency

plugins/aws-serverless/skills/aws-lambda-managed-instances/references/cost-comparison.md

@@ -0,0 +1,72 @@
+# Lambda vs LMI Cost Comparison
+
+## Pricing Components
+
+**Standard Lambda:** $0.20/M requests + $0.0000166667/GB-sec (x86) or $0.0000133334 (ARM). Compute Savings Plans give ~17% discount on duration.
+
+**LMI:** $0.20/M requests + EC2 instance cost (24/7) + 15% management fee on EC2 on-demand price. No per-request duration charge. Savings Plans/RIs discount up to 72% on EC2 compute. 15% fee always on on-demand price.
+
+## Discount Comparison
+
+| Option | Lambda discount | LMI discount |
+|--------|----------------|--------------|
+| On-demand | 0% | 0% |
+| Compute Savings Plan (1yr) | ~17% on duration | ~40-50% on EC2 |
+| Compute Savings Plan (3yr) | ~17% on duration | ~60-72% on EC2 |
+| Reserved Instances (1yr) | N/A | ~40% on EC2 |
+| Reserved Instances (3yr) | N/A | ~60-65% on EC2 |
+
+Compute Savings Plans apply to both Lambda duration AND EC2 instances. One commitment can cover both.
+
+## Calculation Formulas
+
+```
+# Lambda on-demand
+duration_cost = requests × avg_duration_sec × memory_GB × $0.0000166667
+request_cost  = requests × $0.20 / 1,000,000
+total         = duration_cost + request_cost
+
+# Lambda + Savings Plan (17% on duration)
+total         = (duration_cost × 0.83) + request_cost
+
+# LMI on-demand
+ec2_cost      = num_instances × hourly_price × 730
+mgmt_fee      = ec2_cost × 0.15
+total         = ec2_cost + mgmt_fee + request_cost
+
+# LMI + 3yr Savings Plan (65% discount on EC2)
+total         = (ec2_cost × 0.35) + mgmt_fee + request_cost
+```
+
+## Comparison Table Template
+
+Present this for every assessment:
+
+```
+| Component          | Lambda OD | Lambda+SP | LMI OD | LMI+3yr SP |
+|--------------------|-----------|-----------|--------|------------|
+| Requests           | $X        | $X        | $X     | $X         |
+| Duration/compute   | $X        | $X        | $X     | $X         |
+| Management fee     | —         | —         | $X     | $X         |
+| Monthly total      | $X        | $X        | $X     | $X         |
+| Annual total       | $X        | $X        | $X     | $X         |
+| Savings vs Lambda  | baseline  | X%        | X%     | X%         |
+```
+
+## Worked Example
+
+Node.js API, 100 req/s steady (259M req/mo), 200ms avg, 512 MB, x86:
+
+| Scenario | Monthly | Annual | Savings |
+|----------|---------|--------|---------|
+| Lambda on-demand | $484 | $5,808 | baseline |
+| Lambda + 3yr SP | $411 | $4,932 | 15% |
+| LMI on-demand (3× m7i.large) | $288 | $3,456 | 40% |
+| LMI + 3yr SP | $155 | $1,860 | 68% |
+
+## When LMI is NOT Cheaper
+
+- < 50M req/month (fixed 3-instance cost exceeds Lambda)
+- Very short functions (< 100ms duration)
+- Highly bursty, unpredictable traffic
+- Workloads needing scale-to-zero