Archive Assets

Stream a ZIP of the selected designs — the bulk-download path.

A GET with the ids in the query string, not a POST body, because the archive is a pure read and a GET is the only shape the browser can also hand straight to its own downloader (an <a href> or a signed URL later) instead of buffering the whole response in the tab. It is also retryable/resumable-friendly and mirrors the single-file /assets/{id}/download next to it. The per-request file cap keeps the query string well inside the 8 KB request-line limit.

ids: comma-separated asset ids. 400 when the selection is empty or over the file/byte caps (see archive_view.MAX_ARCHIVE_FILES / MAX_ARCHIVE_BYTES), 404 if ANY id isn't the caller's. Unreadable files are skipped and listed in a README inside the archive, never dropped silently.