spof

Run /spof with a description, diagram, or infrastructure-as-code of your system. The analysis covers six categories — infrastructure, network, database, application, third-party vendors, and human/process — and lists each SPOF with its risk level and mitigation.

A SPOF analysis identifies any component without redundancy whose failure would stop the whole system. It includes non-obvious cases like a single vendor, one certificate, or one person holding all deployment knowledge, not just hardware.

Yes, you can attach architecture diagrams, network diagrams, Terraform, docker-compose, or Kubernetes manifests. The skill reads the components and their relationships to identify failure points across all six categories.

Risk equals impact multiplied by likelihood. High impact with medium or high likelihood is critical, while low-impact items stay low priority. This prevents over-prioritizing severe but extremely unlikely failures.

Yes, the Human/Process category covers bus factor risks such as a single person knowing deployment procedures, credentials held by one admin, missing runbooks, and single-approver workflows.

Small systems may not justify full redundancy at every point. The analysis weighs cost-benefit and business impact, recommending proportionate mitigations rather than blanket redundancy for everything.