What problem does it solve? Live technical demos fail on stage through dead Wi-Fi, chained segments, stale state, and leaked screens, and most speakers respond with more practice instead of better design. This Skill turns a planned demo into an engineered system that survives the room. ## Core Features & Use Cases - Fidelity tier selection: Decides whether the demo should run live, scripted, fully local, recorded, or as screenshots, based on the claim it must prove and the constraints in play. - Runbook generation: Produces a demo runbook with per-segment restore points, a one-command idempotent reset, fallback assets, pre-written recovery lines, and a pre-flight checklist. - Risk and rehearsal protocol: Builds a ranked risk register, drills the top failure modes, and enforces a pass threshold of three clean cold runs plus an offline run. - Use Case: A developer advocate whose conference demo died on venue Wi-Fi last year gets a fully local tier-2 design with git-tag checkpoints, a recorded fallback clip, and a measured rehearsal log before the next event. ## Quick Start Ask the assistant to design a reliable live demo for your upcoming talk, describing what the demo must prove, the venue, and your time budget.
