Act as a senior site reliability engineer with a product designer's eye. Set up maintenance mode and a status page for {{business_name}}'s site at {{website_url}}, built with {{tech_stack}}. Read the hosting and deploy setup first, and tell me where each piece can live. Maintenance mode: - A switch, such as an environment variable or edge rule, that serves a static maintenance page from the CDN, so it works even when the app and database are down - Every page returns 503 with a Retry-After header, so search engines don't index the maintenance page or drop rankings - Admins bypass it with a signed cookie or an IP allowlist, to test before reopening - The page says what's happening, when we expect to be back in the visitor's time zone, links to the status page, and gives a way to reach us - A dismissible site-wide banner 48 hours before planned work Status page: - Hosted off our main infrastructure, on a subdomain such as status., with a hosted status tool or a static page - The components people care about, such as website, sign-in, payments and email, each with a state: operational, degraded, partial outage, major outage or maintenance - Incidents as a timeline: what's affected, what we know, when the next update comes, and a short write-up once resolved - Uptime history from real checks by an outside monitor, never a typed-in 99.99% - Subscribe by email or RSS Design: calm and plain, brand color used only for states, every state written in words as well as color, contrast at least 4.5:1, and no horizontal scroll at 360px. No pulsing dots or animated gradients. Finish with how to turn maintenance on and off, a template for incident updates, and the files changed.