The questions that keep arriving.

Not yet, and this is the biggest gap in the product. The designer edits an in-memory document. The schema it will be written to is complete — projects, pages, sections and revisions — but no project loader is wired, so closing the tab loses the document.

It is the next thing being built. Until it lands, copy the markup and the stylesheet out of the code pane; both are plain text and neither needs this app to be useful.

Yes, on any plan, at any time. The code panel is the document rather than a report of it — semantic tags, your class names, and the stylesheet you wrote. What comes out is a page you could have hand-built.

There is no proprietary format to be trapped in and no runtime to include. That is a consequence of the architecture rather than a policy someone could change.

It blocks editing below 1100px. Rather than reflowing into a layout where a drag lands on the wrong element — a design tool that is subtly inaccurate about where you clicked is worse than one that says it needs more room.

Because your stylesheet is injected as authored — not prefixed, not scoped. You are writing the stylesheet your site will ship, and rewriting your selectors would mean what you see is not what you get.

The cost is real: a bare `p { color: red }` also reaches the editor. Every element of the shell carries its own class to limit the blast radius, and the real fix — moving the artboard into an iframe, which also buys a correct viewport for breakpoints — is on the roadmap.

Most of them make the canvas authoritative and generate code out of it. You can read that panel; you cannot meaningfully write in it, because the next drag regenerates over the top. It is an export wearing an editor’s clothes.

Here both views dispatch into one reducer. Dragging a box and typing a number are the same operation expressed twice, which is why the code pane is writable and why round-tripping is a contract rather than a feature.

No. Snapshot-to-CDN with custom domains, previews rendered from the database, branch URLs and immutable deployment snapshots are the chosen architecture, and the schema for all of it exists. None of it is built.

Preview works today and opens the document as a real page in a real tab, with the breakpoint layers emitted as the media queries they represent.

No. Both are copied by value on placement, so there is no live link back. Editing a library block later does not reach into pages that already used it, and a future release changing a preset cannot alter a page you already built.

That is a product decision rather than a missing feature: a live binding means editing one block silently changes every site that placed it, including a client’s published homepage. Placements are recorded so the product can offer an explicit re-sync instead — the safe version of the same power.

No. Comlumen stands alone. If you already use one, your sign-in works here too, because this is a satellite of the same account rather than a separate identity system.

One connection is genuinely useful rather than merely shared: Baselumen’s component library renders live in a panel and imports into your document.

Nothing yet. There is no checkout wired, no billing code in the repo, and no plan grants anything. The tiers are published so the shape can be reviewed; the amounts are not decided, and the page says so at the top.

The intent is that published sites stay up until the end of the period and then serve a holding page, and that documents are never deleted. Downgrading would lock publishing, not your work.

Stated as intent because none of it is implemented. When it is, this answer changes tense.