aquavis12/aws-resilience-hub

v4.0.0MIT

Assess, improve, and govern the resilience of AWS applications from your IDE — next-gen Resilience Hub GenAI failure-mode assessments, dependency discovery, RTO/RPO policy validation, enterprise-grade remediation reporting with resource ARNs and cost validation via AWS Pricing API, Amazon Application Recovery Controller (ARC) failover orchestration, and AWS Well-Architected Reliability-pillar guidance — powered by the managed AWS MCP Server with integrated AWS Documentation, re:Post knowledge, and AWS Pricing.

GitHub Open Website

Install

Kiro → Powers panel → Add Custom Power → Import power from GitHub

https://github.com/aquavis12/power-aws-resilience-hub

Kiro reads plugin.json + mcp.json and installs automatically. The power activates on keywords like resilience, RTO/RPO, failover, disaster recovery, assessment, backup, pricing, cost, report.

Prerequisites: AWS CLI ≥ 2.32.0 · uvx installed · a working aws sts get-caller-identity.

What it does

GenAI failure-mode assessments with automatic dependency discovery, plus RTO/RPO validation against the four DR tiers
ARC failover orchestration — zonal shift, routing controls, Region switch
Enterprise remediation reports — full resource ARNs, per-disruption-layer findings, operational gaps (alarms/SOPs/FIS), console URLs, prioritized roadmap
Cost-quantified fixes — unit rate × quantity = monthly total, validated against the live AWS Price List API
Every finding maps to DR tier, target RTO/RPO, remediating service, Well-Architected best-practice ID, cost, and a docs link
Cross-checks assessment scope against live infrastructure to catch resources the tag filter missed, and flags false-positive "stateless" classifications
Optional CI/CD gate so resilience regressions fail the build

Skills

Skill	What it does
`assess`	Full assessment for one app — validate, discover, assess, report with WAF best-practice IDs
`account-scan`	Account-wide multi-region scan — all apps, unregistered workloads, aggregated posture
`failover`	ARC zonal shift / routing controls / Region switch, with explicit safety confirmation
`report`	Enterprise remediation report from an existing assessment — ARNs, costs, gaps, roadmap

MCP servers

Server	Purpose	Package
`aws-mcp`	All AWS API calls — Resilience Hub, ARC, STS, Backup, DRS	`mcp-proxy-for-aws`
`aws-docs`	Service guides, best practices, console paths	`awslabs.aws-documentation-mcp-server`
`aws-repost`	Community solutions, known issues, real-world patterns	`awslabs.aws-repost-mcp-server`
`aws-pricing`	Live Price List data for remediation cost validation	`awslabs.aws-pricing-mcp-server`

Safety

Read and assess by default. Failover actions move production traffic and always require explicit confirmation. Recovery runs through the data plane, never the control plane (Well-Architected REL11-BP04).

Changelog

v4 — AWS Pricing MCP, enterprise report skill, cost-quantified recommendations, pricing/cost/ARN keywords. v3 — Report skill, docs + re:Post MCP enrichment, cross-infrastructure validation.