What problem does it solve? Incident runbooks and on-call response flows are hard to communicate when ownership, retries, and escalation paths are tangled into a single flowchart. This Skill produces a left-to-right swimlane diagram where each lane is one actor, service, or system, so every step of an incident response shows exactly who owns it and how each path ends. ## Core Features & Use Cases - Ownership swimlanes: Builds one lane per actor or system (User, Support, On-call Engineer, API, Retry Worker, Escalation) with decision diamonds and labelled branch exits. - Separated incident paths: Keeps happy path, retry/backoff, error, dead-letter, fallback, escalation, and resolution branches distinct and traceable, each ending in an explicit terminal state. - Quality-gated output: Runs a lint, score, repair, and validate loop until the drawing scores at least 95 with zero hard blockers, then saves, versions, and exports it. - Use Case: Ask for a swimlane of a checkout 5xx incident and receive a diagram where Support triages, the On-call Engineer classifies transient vs bug, the Retry Worker retries with backoff to recovery or dead-letter, and out-of-SLA cases escalate to L2. ## Quick Start Ask the assistant to draw a troubleshooting swimlane for your incident, naming the lanes, the trigger, the decision points, and the terminal states such as Resolved or Escalated.
