Pinned Tweet
If you’re serious about delivering fast WordPress sites , or just tired of fighting performance without a clear roadmap, this is the course you’ve been waiting for.
Let’s make WordPress fast.
withinwp.com/courses/make-w…
Let’s make WordPress fast.
withinwp.com/courses/make-w…
7
5
22
4,960
Remkus de Vries retweeted
Has your role become your identity? Remkus de Vries asks how AI's influence on WordPress work changes the role you think you have.
remkusdevries.com/wordpress-isnt…
remkusdevries.com/wordpress-isnt…
0
1
0
121
Hard work today! 😁
We’re getting ready for @CloudFestUSA in Miami, taking a tour around the Everglades, with Lasse Hall from @servebolt and @remkusdevries. It’s good to be in the USA again!
0
0
11
466
Remkus de Vries retweeted
You’ve opened a pull request on WordPress’s GitHub. The README calls that repo “just a mirror.” Core still goes through Subversion and Trac. Open source history, read as who could say yes. 🌳 courtneyr.dev/2026/11/05/his… #opensource #WordPress
2
1
5
445
Remkus de Vries retweeted
My thoughts exactly. Shipping your own MCP server is nothing but marketing. The #WordPress MCP Adapter is very capable of serving all abilities. Combine it with @albertwp_ and you'll have the freedom to expose what you want and to who.
If every WordPress plugin ships its own MCP server and connection screen, users have to set up the same site over and over.
Let plugins expose their abilities through a shared connection. We already have enough settings pages in WordPress.
1
1
2
482
Remkus de Vries retweeted
A WordPress setup screen can return HTTP 200 while leaving the site open to takeover.
Scanfully 2.5 treats a detected setup or installer screen as downtime. Once confirmed, its urgent alert bypasses quiet hours and the digest. An active snooze still applies.
Scanfully 2.5 treats a detected setup or installer screen as downtime. Once confirmed, its urgent alert bypasses quiet hours and the digest. An active snooze still applies.
1
1
0
142
Remkus de Vries retweeted
WordPress Studio’s rebuilt workspace, wrapido turning REST routes into CLI commands, and It Looks Done’s fictional launch day.
This week’s Within WordPress:
withinwp.com/2026/woocommer…
This week’s Within WordPress:
withinwp.com/2026/woocommer…
0
1
0
115
Remkus de Vries retweeted
Monitoring can fail without a sound. The data just stops updating.
If the Scanfully plugin goes quiet, you now hear why and what to do.
Or ask your AI app:
"Which of my sites stopped reporting, and why?"
Works in Claude, ChatGPT, Cursor and VS Code.
If the Scanfully plugin goes quiet, you now hear why and what to do.
Or ask your AI app:
"Which of my sites stopped reporting, and why?"
Works in Claude, ChatGPT, Cursor and VS Code.
1
1
2
144
This is a must have for anyone running a WooCommerce store. And especially with the Black Friday cycle we find ourselves in now.!
Built something for @Scanfullyapp that I wished I had years ago. Fully automated @WooCommerce purchase flow checks. From product page to checkout. Never wonder if your checkout still works.
2
1
3
464
Read more about it in our latest release: scanfully.com/introducing-sc…
1
0
0
120
Remkus de Vries retweeted
“Your site is down” leaves you with the diagnosis.
Scanfully 2.5 names detected WordPress errors, database errors, timeouts, and connection failures in the alert, with a next step.
When PHP exposes the error, it can also name the plugin or theme.
Scanfully 2.5 names detected WordPress errors, database errors, timeouts, and connection failures in the alert, with a next step.
When PHP exposes the error, it can also name the plugin or theme.
1
1
0
115
Remkus de Vries retweeted
This WordPress page loads fine. It also lets anyone take over the site.
Lose wp-config.php in a failed deploy or migration, and WordPress opens its setup screen. A monitor that only reads the status code sees a healthy site.
Scanfully sends an urgent alert, past quiet hours.
Lose wp-config.php in a failed deploy or migration, and WordPress opens its setup screen. A monitor that only reads the status code sees a healthy site.
Scanfully sends an urgent alert, past quiet hours.
6
1
4
225
Remkus de Vries retweeted
1/12 With 60K+ plugins available, WordPress pretty much allows you to add any functionality you like.
Versatile, but not smart to do just anything in WordPress. Here's 2 things you should never (want) to do in WordPress.
🧶
Versatile, but not smart to do just anything in WordPress. Here's 2 things you should never (want) to do in WordPress.
🧶
19
48
202
Remkus de Vries retweeted
Ymir manages AWS infrastructure in an account you control for WordPress, WooCommerce and Laravel. Your developers still maintain application code, dependencies and configuration. Carl Alexander's site explains it.
ymirapp.com
ymirapp.com
1
1
0
99
Remkus de Vries retweeted
Modern Font Stacks groups ready-to-copy system font stacks by style, with previews. The exact font varies by operating system, but you avoid another web-font download. Handy resource.
modernfontstacks.com
modernfontstacks.com
0
1
1
147
Remkus de Vries retweeted
Matt Schwartz compared 12 WooCommerce hosting configurations with roughly 120,000 orders, 15,000 subscriptions and a 5GB database. Servebolt led the combined score; BigScoots was fastest in admin and order search.
Read the method: most tests were July 2025; Kinsta was July 2026 with newer software. Hosts used recommended setups. Best-of-run averages and extra weight for order search shape the scores.
The article is on Servebolt's site. Servebolt supplied its test environment and sponsored the Kinsta test. These conditions belong alongside the results.
servebolt.com/articles/best-…
Read the method: most tests were July 2025; Kinsta was July 2026 with newer software. Hosts used recommended setups. Best-of-run averages and extra weight for order search shape the scores.
The article is on Servebolt's site. Servebolt supplied its test environment and sponsored the Kinsta test. These conditions belong alongside the results.
servebolt.com/articles/best-…
0
1
0
118
Second week in a row that I find myself on train. Again to Schiphol. Together with team @greyd_suite I’m going to CloudFest Americas in Miami.
Who will I see there?
Who will I see there?
4
0
32
845
Remkus de Vries retweeted
🔥 Greyd Conversations #21 is out! In this episode, "The Hosting Problem Nobody Solved for Marketing Teams" our COO Sandra Kurze met with our CEO Mark Weisbrod and Servebolt CEO Lasse Hall at the first PressConf Europe in Barcelona to talk about why marketing teams at franchises, prises often wait weeks to get new websites live – even with the best teams and website creation software in place.
They discuss the hidden business cost of slow rollouts, why hosting has never truly caught up with the needs of non-technical users, and how the new Greyd and Servebolt partnership closes that gap.
You’ll gain practical insights into scaling multisite rollouts, one-click site creation, and the business case behind building pipelines that remove the bottleneck between marketing ambition and site launch – recorded live at PressConf Europe in Barcelona.
Some takeaways from their conversation:
✅ A week of waiting is a week your campaign can’t bring in leads. The launch date moves, and so does the start of your sales pipeline.
✅ Your hosting invoice misses a big part of the cost. Every handoff takes time from the people coordinating the rollout.
✅ Marketing teams can own the launch date and still depend on IT to get the site set up. That’s where a ready-to-go campaign can sit for weeks.
✅ Faster site creation only gets you so far. If every new site still needs a separate hosting setup, you’ve sped up one step and left the wait in place.
✅ The same bottleneck shows up in franchises, universities, and enterprises. When several teams have to coordinate every launch, adding more sites means repeating that work.
Watch the episode here: eu1.hubs.ly/H0zDnNm0
They discuss the hidden business cost of slow rollouts, why hosting has never truly caught up with the needs of non-technical users, and how the new Greyd and Servebolt partnership closes that gap.
You’ll gain practical insights into scaling multisite rollouts, one-click site creation, and the business case behind building pipelines that remove the bottleneck between marketing ambition and site launch – recorded live at PressConf Europe in Barcelona.
Some takeaways from their conversation:
✅ A week of waiting is a week your campaign can’t bring in leads. The launch date moves, and so does the start of your sales pipeline.
✅ Your hosting invoice misses a big part of the cost. Every handoff takes time from the people coordinating the rollout.
✅ Marketing teams can own the launch date and still depend on IT to get the site set up. That’s where a ready-to-go campaign can sit for weeks.
✅ Faster site creation only gets you so far. If every new site still needs a separate hosting setup, you’ve sped up one step and left the wait in place.
✅ The same bottleneck shows up in franchises, universities, and enterprises. When several teams have to coordinate every launch, adding more sites means repeating that work.
Watch the episode here: eu1.hubs.ly/H0zDnNm0
0
1
2
154
Remkus de Vries retweeted
A scan could stay marked as running for up to two hours if it was in progress while we updated Scanfully.
We fixed that in 2.5.0, so this update-related status no longer leaves you waiting.
docs.scanfully.com/changelog/#250…
We fixed that in 2.5.0, so this update-related status no longer leaves you waiting.
docs.scanfully.com/changelog/#250…
1
1
0
116
Remkus de Vries retweeted
This was very interesting to build. Most of the times it's as simple as 2XX code = site works. But there's some weird, and potentionally dangerous, edge cases.
Getting a “your site is down” alert usually means opening the site and the hosting panel to work out what you're dealing with. Is WordPress throwing an error, or do you need to get the host involved?
We've just updated Scanfully to read what a failing page is actually showing ation, with suggested next steps, in the alert.
So you'll see whether it's a WordPress critical error, a database connection failure, or a PHP fatal error. Where the page exposes the PHP error details, you get the message, file, and plugin or theme involved. Updates leading up to the downtime are there too, so you can check what changed.
We've also added detection for something a status-code check can miss entirely: the WordPress setup screen.
If wp-config.php goes missing or the database gets emptied, WordPress can start offering to set itself up again. The page loads, the monitor sees a successful response, and your site can be open to takeover. Once Scanfully confirms it, the urgent alert goes out, even during quiet hours and without waiting for the daily digest.
Then there's the monitoring plugin itself. If it stops reporting, your site health, update, vulnerability, and activity data quietly gets stale. Scanfully now flags that too, identifies the reason where it can, and tells you what to check. That might be a restored backup, scheduled tasks that aren't running, or a firewall blocking the connection.
You can access this context through our MCP server as well. Connect Claude, ChatGPT, Cursor, or VS Code and ask “Why was the shop down last night, and what should I do?” or “Which of my sites stopped reporting, and why?”
A few other changes in this release:
- Downtime is counted from the first failed check across alerts, the digest, and uptime screens.
- WordPress maintenance mode gets 10 minutes, so a normal update doesn't trigger downtime.
- Timeout alerts arrive about 2 minutes sooner.
You can try Scanfully free for 14 days. No credit card needed
app.scanfully.com/sign-up
1
2
2
172