Payload Portable Plugin

Export all configured Payload collections and globals to a single JSON file and import them again through actions in the admin panel.

Usage

import { postgresAdapter } from "@payloadcms/db-postgres"
import { payloadPortablePlugin } from "@mvriu5/payload-portable-plugin"

export default buildConfig({
    db: postgresAdapter({
        allowIDOnCreate: true,
        pool: {
            connectionString: process.env.DATABASE_URL!,
        },
    }),
    plugins: [payloadPortablePlugin({ importMode: "merge" })],
})

The admin header provides Import and Export actions. The same actions appear next to Create New in each included collection list and operate only on that collection.

Documents are matched by ID. In "merge" mode, a missing ID falls back to the collection's populated unique fields. The resulting source-to-target ID map is applied to relationship, rich-text relationship, and upload fields before documents are written. No import mode deletes documents. Localized content is transferred for every configured locale.

Empty non-default locale variants are omitted. When localized required fields exist, at least one of them must contain a value; generated values such as a localized slug alone do not mark the locale as translated. The import therefore leaves untranslated locales untouched instead of triggering required-field validation.

Authentication collections are always excluded. Upload collection metadata is included so existing target media can be matched by filename and referenced with its target ID. Binary files are not embedded, so an upload without a matching target file cannot be created from the archive.

By default, all endpoints require an authenticated user. Every individual read and write operation also enforces the access-control rules of its collection or global.

allowIDOnCreate: true is required on the database adapter to create missing documents with their original IDs. Without this option, documents matched by ID or a unique field can still be updated, while unmatched documents are skipped and reported.

Import mode

The required importMode plugin option controls how documents are imported:

• "merge" creates missing documents and updates existing documents
• "add" creates only missing documents and leaves existing documents unchanged
• "replace" updates only existing documents and skips missing documents

Globals are updated in "merge" and "replace" modes. They are skipped in "add" mode because globals always exist.

Options

payloadPortablePlugin({
    access: ({ req }) => req.user?.roles?.includes("admin") === true,
    batchSize: 250,
    excludeCollections: ["payload-preferences"],
    excludeGlobals: ["internal-settings"],
    importMode: "merge",
    placeholderData: {
        media: {
            alt: "Missing imported image",
        },
    },
})

• importMode: required import mode: "merge", "add", or "replace"
• access: additional authorization check; authenticated users are allowed by default
• batchSize: export page size; defaults to 100 and is limited to 1000
• excludeCollections / excludeGlobals: additional slugs to skip during both import and export
• placeholderData: additional data keyed by upload collection slug for generated placeholder documents
• disabled: disables the admin actions and endpoints

Notes

The archive contains collection documents, upload metadata, and global data, including relationship and upload references. Binary files and secret authentication data hidden by Payload are not embedded.

Hooks, validation, and access control run normally during imports. Schema mismatches are therefore included in the import report for each affected document.

If an import contains errors, the admin UI automatically downloads a sanitized JSON error report. Repeated errors are grouped by entity and error code, with affected IDs, locales, and a suggested resolution. Full technical errors remain available in the Payload server log; SQL queries, local file paths, and stack traces are not exposed in the downloaded report.

Relationship failures are retried automatically after the initial import pass. The plugin continues retrying while at least one queued item succeeds, allowing documents that were imported out of dependency order to resolve later. Unresolvable or circular relationships remain in the final error report.

Missing upload relations are resolved before documents are written. When exported upload metadata matches an existing target filename, upload fields are rewritten to that target media ID. Otherwise, required image fields receive a shared 1×1 PNG and required video fields receive a small MP4 placeholder; optional missing upload relations are removed. The plugin reuses one placeholder per upload collection and automatically fills required text and textarea fields with Import placeholder. Use placeholderData when an upload collection has additional required fields or needs custom values. These replacements are reported as grouped warnings so editors can replace placeholders after the import.