The total bill to make everything resilient came to $11.06 a month. The uncomfortable part was working out why I hadn't already paid it.
Metric | Value |
|---|
Apps assessed | 4 |
Average resiliency score | 0.4 / 1.0 |
Components with a Region gap | 10 of 16 |
Monthly cost to fix all of it | $11.06 |
Why it needed to live in the IDE
Resilience Hub is a good service. The next-generation version that went GA in May 2026 does GenAI failure-mode assessments, dependency discovery, RTO/RPO validation against live infrastructure. I'd used it maybe three times.
The problem was never the service. It was where the service lives. Leave your editor, open the console, create a policy, register an app, import resources, publish a version, run an assessment, wait, read the findings, then translate them back into the CloudFormation sitting in the tab you left twenty minutes ago. Every one of those steps is a place to stop.
So the findings stayed in the console and my IaC stayed unresilient.
What the power is
Four MCP servers, each covering a different part of one question. Is this workload resilient, and what does fixing it cost?
Server | What it brings | Package |
|---|
aws-mcp | Every AWS API operation: Resilience Hub, ARC, STS, Backup, DRS | mcp-proxy-for-aws |
aws-docs | Remediation guidance, console paths, service references | awslabs.aws-documentation-mcp-server |
aws-repost | Community solutions and known issues from AWS re:Post | awslabs.aws-repost-mcp-server |
aws-pricing | Live Price List API data for costing every recommendation | awslabs.aws-pricing-mcp-server |
The split matters. aws-mcp tells me a bucket has no cross-Region replication. aws-docs points at the Config rule and the CloudFormation property that fix it, aws-repost surfaces what broke for the people who tried it before me, and aws-pricing puts the number at 43 cents a month. Any one of those on its own just leaves me with homework.
Plus four skills ( assess , account-scan , failover , report ) and three steering files. Skills are the part people underrate. An MCP server gives the agent capability. A skill gives it a procedure.
What it found
The account scan turned up one registered app and three CloudFormation workloads that had never been near Resilience Hub. Resilience Hub only assesses what you register, and what you register is what you were already worried about. The ones that quietly matter never get added.
Across all four apps:
- Average resiliency score 0.4 out of 1.0, zero policy compliant
- 10 of 16 components with no regional recovery path
- 86 missing CloudWatch alarms, 34 missing SOPs
- Every stateful data store, S3 and DynamoDB alike, with zero cross-Region DR
Read the full version
The complete write-up has the per-component assessment breakdown, the eight systemic root causes that repeated across all four apps, the full pricing table, the screenshots, and the five things I got wrong building this:
Loading article