How the ElastiCache skill works
The ElastiCache skill routes your request to the appropriate workflow area based on your intent. It loads only the reference material needed for the current task. The following table describes the workflow areas included in the skill.
| Workflow area | Purpose |
|---|---|
| Solution fit | Gathers workload requirements and recommends whether ElastiCache is the right fit for your application. |
| Create and connect | Helps create caches, choose engines, configure networking, set up authentication, and connect applications. |
| Application patterns | Helps choose key schemas, TTLs, invalidation strategies, and Redis or Valkey data structures. |
| AI and vector workloads | Helps implement semantic caching, agent memory, RAG, recommendation, personalization, and vector search patterns. |
| Operate and observe | Helps configure dashboards, alarms, logs, troubleshooting, cost reporting, and security audits. |
| Migration | Helps plan and run migrations from self-managed Redis, Redis OSS, or another ElastiCache deployment model. |
Setup and networking
The skill helps agents identify the appropriate deployment configuration for a workload before generating setup guidance or commands. It guides agents to consider engine choice, serverless or node-based deployment, feature compatibility, and other requirements. The skill then guides agents through networking setup, including VPC placement, subnet and security group configuration, and application connectivity.
For local development, the skill guides agents to use supported access patterns such as a jump host or AWS Systems Manager Session Manager tunnel, instead of assuming direct access to an ElastiCache endpoint from a local machine.
The skill stores ElastiCache-specific project context in a state file at
.elasticache/requirements.json. This file helps the agent avoid asking for the
same information repeatedly. It can include values such as engine, AWS Region, and cache
endpoints. If this file exists, the skill reads it before asking new questions.
Cost estimation
The skill can help agents provide preliminary pricing guidance for ElastiCache workloads. The skill includes a pricing calculator to estimate monthly costs from workload inputs such as data size, request volume, command types, and payload size.
AWS credentials are not required to generate pricing estimates. The pricing calculator uses publicly available AWS pricing data, so you can estimate serverless and node-based costs before provisioning resources or connecting an AWS account. The pricing estimates are for planning purposes only and can vary from actual costs. Before deployment, confirm current pricing on the ElastiCache pricing page.
Monitoring and troubleshooting
The monitoring workflow helps diagnose operational issues by starting with relevant metrics and then recommending the smallest effective change. The skill can use bundled scripts to collect metrics, generate dashboards, run security audits, and support cost analysis.
Use the ElastiCache skill to investigate performance, availability, and cost issues such as the following:
High latency
Low cache hit rates
CPU or memory pressure
Replication lag
Connection spikes
Hot keys and big keys
Shard imbalance
Slow commands
Unexpected cost increases
CloudWatch alarm configuration
Log delivery
GenAI and vector workload guidance
The skill helps agents provide guidance for AI application patterns such as semantic caching, AI agent memory, retrieval-augmented generation (RAG), recommendation systems, and vector search. The skill helps agents identify the user's intent, select the appropriate workflow, and generate implementations that are aligned with ElastiCache capabilities.
For these workloads, the skill provides detailed guidance for embedding storage, key design, TTL strategy, similarity threshold tuning, framework integrations, and operational considerations.