Resilience Hub Failure Mode Assessment

Failure Mode Assessment

Overview

Domain expertise for running Resilience Hub v2 failure mode assessments, interpreting findings, triaging by severity and achievability, and driving remediation.

The AWS MCP server is recommended for executing this skill’s AWS API calls, but it is not required — all operations also work with the AWS CLI directly.

Guardrail — where this skill’s own files live (MCP vs local install)

Before reading a reference file, determine how this skill was loaded:

• Loaded via the AWS MCP retrieve_skill tool: the skill’s reference files are not on the local filesystem. Fetch each one through retrieve_skill with the file parameter (e.g. file="references/assessment-workflow.md") — do NOT file_read these paths locally or search the filesystem for them.
• Installed locally (e.g. .kiro/skills/resilience-hub-failure-mode-assessment/ or ~/.claude/skills/resilience-hub-failure-mode-assessment/): read reference files from the local skill directory using the relative paths shown here.

This applies only to the skill’s own reference files; always read and write user or session data in the working directory, never through retrieve_skill.

Run and interpret assessments

To run assessments and triage findings, follow the procedure exactly. See references/assessment-workflow.md (opens in a new tab).

Troubleshooting

Assessment fails with INVALID_PERMISSIONS

The service’s permission model (invokerRoleName / crossAccountRoles) doesn’t have access to the resources. Verify the invoker role (and any cross-account roles) can describe resources in all configured regions.

Too many findings — where to start?

Prioritize by finding severity, highest first (HIGH, then MEDIUM, then LOW). For HIGH-severity findings, check the service’s achievability for the relevant policy component (from get-service / list-failure-mode-assessments): NOT_ACHIEVABLE means the architecture must change before testing; ACHIEVABLE means validate the fix with an FIS experiment. MEDIUM findings: plan remediation this sprint; LOW findings: track but don’t block (see the priority matrix in references/assessment-workflow.md (opens in a new tab) Step 5).

AI-generated service functions are wrong

Update them: aws resiliencehubv2 update-service-function to rename or change criticality (there is no service-function “type” parameter). Reassign resources by calling create-service-function-resources with the desired resource set (see references/assessment-workflow.md (opens in a new tab) for the service-function operations).

Security Considerations

• Least privilege: the invoker role should be scoped to read-only discovery of only the resource types in the service’s input sources; avoid granting access beyond what assessment needs.
• Encryption & access control: recommend that S3 buckets used for report output have server-side encryption (SSE-S3 or SSE-KMS) and block public access — assessment reports can contain sensitive architectural detail. If a bucket policy grants the Resilience Hub service principal write access, scope it with aws:SourceArn / aws:SourceAccount condition keys to prevent confused-deputy writes.
• Further reading: see Security in AWS Resilience Hub (opens in a new tab) and the AWS Well-Architected Security Pillar (opens in a new tab) for securing assessment outputs and IAM configurations.