Data import and export

Download a JSON copy of your content, users, audience, settings and email data, and a ZIP of your media files, then load them back into the same install or a new one.

What it is

The Data admin menu has four screens:

Everything runs in the admin's browser, with the admin's own permissions. There are no Cloud Functions involved.

For creating content entries from a spreadsheet, use the content type's Bulk Import instead (Content). For contacts from a CSV, see Audience.

Turn it on or off

Data import and export is a switchable feature with the id data. It is on by default and needs no other feature. To turn it off, list it in src/custom/features.ts:

export const CUSTOM_FEATURES: FeatureChoice = {
    off: ['data'],
};

When it is off, the Data menu goes and its addresses show the not-found page. Nothing is stored by this feature, so nothing is left behind. See Feature ids.

Set it up

Nothing to set up. The screens need an admin account and the deployed security rules.

Use it

Export data

1. Open Data > Export Data.

2. Pick a preset (Everything, All Content, All Settings) or tick collections by group. Picking a content type takes its drafts, its published entries and its tags together.

3. Export. The browser downloads arccms-export-{date}.json.

Collections of a feature that is off are not offered. Nothing else is exported: not Authors, not content translations or version history, not payments, feedback or notification data.

The file format

{
  "version": "1.0",
  "exportedAt": "2026-09-30T10:00:00.000Z",
  "collections": {
    "ContentTypes": { "articles": { "name": "Articles", "slug": "articles" } },
    "Waitlists/waitlist/users": { "abc123": { "email": "a@example.com" } }
  },
  "metadata": { "totalDocuments": 2, "collectionSummary": [ { "name": "ContentTypes", "count": 1 } ] }
}

Keys under collections are collection paths, then document ids. Firestore timestamps, document references and geo points are written as objects with a __type of timestamp, ref or geopoint, and turned back into those types on import.

Import data

1. Open Data > Import Data and choose an export file. It is checked for a version and a collections object; unknown collections are flagged but can still be imported.

2. Tick the collections to import.

3. Choose what happens to a document that already exists:

   Mode	Existing document
   Skip existing (default)	Left as it is
   Overwrite	Replaced by the file's version
   Merge	The file's fields are written over it; other fields stay

4. Import. Document ids are kept. Collections go in a fixed order so that what others refer to comes first: content types and settings, then tags, users, drafts and published entries, media records and email templates, then signup forms, audience, broadcasts and email logs. Writes go in batches of 500. A content type whose slug another type on the site (or earlier in the file) already uses is not imported, and the result lists it as an error: the same type, with the same document id, is imported as usual. A type already on the site, by document id, whose slug in the file differs (or is missing, when existing documents are overwritten) is not imported either while entries, drafts or tags are stored under its current slug, since they would be left behind; the edit form locks the slug for the same reason. To restore a type over one the setup wizard created, such as Articles, delete the wizard's type first, then import.

Export files

Data > Export Files lists the media library: its images, then its files (PDFs and other documents, shown with their first page or an icon). Select items and download arccms-media-{date}.zip: one file per item (the largest stored size of an image) and media-manifest.json with each item's name, URL, Storage path and upload time, its title, alt text and notes when it has them, and "collection": "mediaFiles" for a file.

Import files

Data > Import Files has two modes:

With Update media collection metadata ticked (the default), each uploaded file also gets a new media record (a mediaFiles record for a document), with a new id. Unzip the export first: the screen takes files, not the ZIP. Imported files are stored as they are, at one size.

Extend it

The exported collections are listed in KNOWN_COLLECTIONS in src/app/pages/admin/(data)/data-constants.ts, a core file. An app's own collections are not exported; back them up with your own tooling, or say so as a core feature request (Keep core generic).

What it stores

Nothing of its own. Export reads the collections above; import writes them, plus Storage files and media records (downloadURL, name, storagePath, uploadTime) for imported files. There are no Cloud Functions. Collection details are in Data model.

Who can do what

The Data screens are admin only. Every read and write is the admin's own, so each collection's security rules apply as usual: an admin can read and write everything offered except Contacts and Suppression, which admins can read but only Cloud Functions write. Storage uploads need an admin (or the reserved editor role). See Security rules.

Troubleshooting