Definition: A kill switch is a control that instantly makes an internal application unreachable for all users — without deleting it, without a deploy, and without touching infrastructure. It is the "off now, investigate after" button.
The concept matters most for AI-built tools. A vibe-coded app is typically written quickly, reviewed lightly, and connected to real systems — a CRM API, a database, a spreadsheet of customer data. When something goes wrong with a tool like that, the time between "we should turn this off" and "it is off" is pure exposure.
What a real kill switch requires
Not every hosting setup can actually deliver one:
- Enforcement in front of the app, not inside it. If the off-check lives in the app's own code, a buggy or compromised app can't be trusted to enforce it. VibeHub's gateway enforces the switch before a request ever reaches the app.
- Instant effect for everyone. Not "propagates in 24 hours", not "everyone please stop using it."
- Reversibility. State — versions, secrets, audit history — survives the off state, so investigation and recovery remain possible.
- Authorization and audit. Flipping it is restricted to workspace admins and recorded: who turned what off, when.
Kill switch vs rollback
They answer different failures. Rollback answers "the new version is broken — restore the last good one." A kill switch answers "this app should not be running at all right now." Mature setups need both, plus the audit trail to reconstruct what happened.
A kill switch is one of the four lifecycle controls of an internal tools control plane, alongside versioned rollback, audit logging, and transferable ownership — all standard on every app deployed to VibeHub.