FAQ
The honest answers — including what RIFT deliberately doesn't do.
Is there a WYSIWYG editor or page builder?
No, and there won't be. RIFT is built for teams that write their own HTML or generate it with AI tools, then want a precise content framework — folders, workflow, versioning, publishing — underneath it. If your authors need drag-and-drop blocks, RIFT is the wrong tool, and we'd rather tell you that now.
Does RIFT modify my code?
No. No sanitization, no escaping, no rewriting. Custom CSS, JavaScript, embeds, and web components ship exactly as written. Safety comes from versioning, review workflow, and the audit log — not from filtering your markup.
Where does my published site run?
Anywhere that serves static files from a git branch: Vercel, Netlify, Cloudflare Pages, nginx, S3-style hosting. The published site has zero runtime dependency on RIFT — the CMS being down never takes a page offline.
Where does RIFT itself run?
Your instance, your infrastructure. RIFT deploys as containers with Postgres and Redis; a provisioning tool stands up a fully isolated instance on Google Cloud in one pass. Nothing about your content transits a shared multi-tenant service.
What about the AI features and data residency?
The AI provider is configured per instance via environment settings only — deliberately not editable in the UI — so an operator decides exactly which model endpoint content is allowed to reach, including fully self-hosted OpenAI-compatible endpoints inside your own network. AI features degrade gracefully to "off" when no provider is configured.
Can I connect my own AI assistant?
Yes — RIFT ships an MCP server, so Claude, Cursor, or any MCP-capable agent can read your content and propose changes under a least-privilege agent key. Nothing an agent does reaches the live site without a human approving the change-set. The MCP guide covers setup in two steps.
Is it accessible / Section 508 ready?
Accessibility scanning is built into the platform: pages are scanned against WCAG criteria with reports and per-item findings, and version history preserves the record of every change to every policy page. The scanning runs inside your instance, not through a third-party service.
Can I import existing content?
Yes — three ways. Word documents convert to clean semantic HTML on import. Rich paste transforms Word/Google Docs clipboard content on the fly. And for whole sites, the migration pipeline crawls, classifies, and moves your existing site with human review at each step.
What does it cost?
We're in design-partner stage: a small number of teams get white-glove onboarding — including an operated migration — in exchange for shaping the roadmap. Pricing beyond that isn't public yet. If the fit sounds right, ask us directly.