Moving sites around has been part of every WordPress project I’ve worked on. You build the site locally, and at some point it has to move to production, usually by way of a staging environment. Once it’s live, you’ll probably add a few feature branches with their own development sites as well. At the end of all that, you’ll want a copy back on your own computer to try out updates and changes without breaking anything for your visitors.
WordPress has never had a built-in way to do any of this. Tools → Export only covers your content, so for the full site – themes, plugins, uploads and settings – you’ve had to reach for a third-party migration plugin, often with the useful parts behind a paid tier. Even though I’ve had good experiences with many of them, site migrations are a big task to entrust to a plugin managed by companies or individuals you don’t know.
Reprint fills that gap. It’s a free, open-source tool from the WordPress project itself, built as part of the Data Liberation work, and it copies a site’s files and database directly to the computer or server you choose. In this tutorial, we’ll use it to make a local copy of a live site, and then go over how the same steps move a site to production or a new host.
A word of warning before we start, though: Reprint is run from the terminal, which makes it a lot more technical than a regular plugin. Each step below is a command you can copy and paste, but if you’re not comfortable working in the terminal, a third-party migration plugin with a regular settings screen might still be the better choice.
How Reprint works
Reprint has two parts:
- The Reprint plugin, which you install on the site you want to copy. It doesn’t do anything on its own – it just lets a client you’ve authorized download the site’s files and database.
- The Reprint client, a single file called
reprint.pharthat you run in the terminal on the computer or server that should receive the copy.
The client pulls the site over in pieces, so the host never has to build one huge backup archive. If the connection drops, you can run the same command again and it picks up where it left off. It also replaces the old site address with the new one throughout the database, including in blocks, shortcodes and serialized data. The files and database go straight from your site to your computer, without passing through anyone else’s servers on the way. Given what I said about trusting third parties above, that’s a big part of why I like it.
Here’s what you need:
- Administrator access to the site you’re copying, and an HTTPS address for it.
- A terminal on macOS or Linux, or WSL on Windows.
- PHP 7.4 or newer on the command line, on the computer receiving the copy. Run
php --versionto check. Macs no longer come with PHP, so if you get “command not found”, install it first. With Homebrew, that’sbrew install php.
You don’t need SSH access to the source site, an account anywhere or a separate database server. The copy can use SQLite.
Reprint is still new, at version 0.11 as I write this, so expect some of the details below to change as it matures.
Step 1: Install the Reprint plugin
On the site you want to copy, go to Plugins → Add New Plugin, search for “Reprint”, then install and activate it. You’ll find its settings under Tools → Reprint Server. Leave that screen open for now.

Step 2: Download the client
Open the terminal on your computer, create a folder for the copy and download the client into it:
mkdir my-site-copycd my-site-copycurl -fL https://plugins.svn.wordpress.org/reprint/downloads/reprint.phar -o reprint.pharphp reprint.phar --version
If everything works, the last command prints a version number.
Step 3: Authorize the client
The plugin won’t hand over anything until you’ve told it which client to trust. Generate a key, replacing https://example.com with your site’s address:
php reprint.phar keygen https://example.com --state-dir=./state --out=./reprint-key.pem
The command prints a long Migration Token. Copy it, paste it into the field under Tools → Reprint Server and click Authorize client. The screen changes to “Client authorized”.
The private key stays in reprint-key.pem on your computer, and the token only allows downloads. It can’t change anything on your site.
Step 4: Copy the site
Now for the actual copy. Run this in the same folder, again with your own site address:
php reprint.phar pull https://example.com \ --private-key-path=./reprint-key.pem \ --state-dir=./state --fs-root=./files \ --target-engine=sqlite --new-site-url=http://localhost:8881
Reprint connects to the site, downloads the files and the database, imports the database into SQLite, swaps the site address for http://localhost:8881 and starts a local server. The terminal shows each step as it goes:
Pulling example.com ✓ Connecting — WordPress 7.1 ✓ Scanned — 5,476 entries ✓ Pulling files — 5,476 changed files pulled ✓ Pulling database — 851.0 KB ✓ Applying database — 41 statements ✓ Preparing runtime ✓ Ready at http://localhost:8881 Press Ctrl-C to stop.
On my test site, this took about ten seconds. A bigger site takes longer, so leave the terminal open until it’s done.
Step 5: Check the copy
Before you start clicking around, keep in mind that the copy is still connected to your live services. Email, payment gateways and scheduled tasks use the same settings as the live site. Turn them off or switch them to test mode first, or your copy might send emails to real customers.
Open http://localhost:8881 in your browser. Click around a bit: the pages should load, the images should show, and links should stay on the copy rather than lead back to the live site. You can log in at /wp-admin/ with the same username and password as on the live site.

Reopening and refreshing the copy
To stop the local server, press Ctrl-C in the terminal. The copy stays in the folder, and you can start it again later with:
bash state/runtime/start.sh
Use this whenever you want to get back to your copy. Don’t run the pull command again for that, because pull replaces the copy’s database with a fresh one from the live site. Anything you’ve changed in the copy’s posts, pages or settings is lost.
When you do want to bring the copy up to date with the live site, pull is the right command. It’s smart about files: the second time I ran it, it only downloaded the two files that had changed.
Removing access when you’re done
The authorized client can download your whole site, database included, so don’t leave it connected longer than you need. Once you’re done copying, go to Tools → Reprint Server on the live site, open Manage client access and remove the client’s Migration Token. It’s listed by the same ID that the keygen command printed. If you won’t be using Reprint again, deactivate and delete the plugin too.
Moving a site to production or a new host
Moving a site works the same way, except that you run the client on the server the site is moving to, and it pulls the site from the old one. Two things need to be in place:
- The site you’re moving from has to be online, with an HTTPS address. Reprint can’t reach a development site that only runs on your own computer, so put it on a staging address first.
- You need command-line PHP on the new server. In practice, that usually means SSH access.
The pull command is the same as above, with a few changes: point it at a MySQL database on the new server, use your real domain as --new-site-url and skip the local development server. You then point the web server at the copied files and check the site before switching the domain over.
The Move to another host guide in the Reprint handbook covers each step. While the copy runs, the old site stays online. So pause new posts, orders and uploads before you make the final copy, or they’ll be left behind.
Things to watch out for
- Reprint is copied too. The plugin comes along with everything else and stays active in the copy, so deactivate it there.
- Some host-specific plugins are left out. Reprint skips the plugins and must-use plugins that hosts like WP Engine, Kinsta and Hostinger add, since they don’t work anywhere else. Features that depend on them won’t be in the copy.
- If the copy fails, check whether a firewall is blocking it. A Cloudflare browser challenge or a security plugin that limits requests will stop the client. Some plugins also need MySQL rather than SQLite. The handbook’s troubleshooting guide covers both.
If you get stuck, you can ask questions in the plugin’s support forum.






