8 min read
The Trigger.dev leak: rotating secrets without downtime
How to identify potential leaks, rotate environment variables in the right order and look for unauthorised use without unnecessary disruption.
Read the article“You have paying customers”
It works, people pay for it, and every change feels like pulling a block out of a tower. That's not cowardice — it's that nobody knows what will break, because nothing tells you. We make it safe to change again, without pausing sales while we do it.
No contract. You talk to the person who builds it.
Why it feels like this
A product with real customers in it becomes dangerous to change for five reasons, and none of them is about what the code looks like.
So every change is tested by your customers, in the live product, and you find out through support.
A few steps in the right order that only you know, and no way back when the fifth one goes wrong.
You change the price in one place and the invoice goes wrong in another. Nobody designed that connection — it just grew.
No test environment, so the only place to try something is where the customers are.
Which is manageable until you're ill, on holiday, or want to sell the company.
Why it feels like this
A product with real customers in it becomes dangerous to change for five reasons, and none of them is about what the code looks like.
So every change is tested by your customers, in the live product, and you find out through support.
A few steps in the right order that only you know, and no way back when the fifth one goes wrong.
You change the price in one place and the invoice goes wrong in another. Nobody designed that connection — it just grew.
No test environment, so the only place to try something is where the customers are.
Which is manageable until you're ill, on holiday, or want to sell the company.
How we do it
First, a safety net
Tests around whatever costs money when it goes wrong: payments, sign-in, anything that counts. Not around everything — around the expensive parts.
Shipping an update becomes a button
One click, with the same result every time, and a way back that's just as simple.
You get a copy to try things in
The same as the live product, with made-up data. There, you and I can be wrong without any customer noticing.
Then we change what you wanted to change
The things that sat untouched for six months because nobody dared. Now a mistake costs a rollback rather than a weekend.
How we do it
First, a safety net
Tests around whatever costs money when it goes wrong: payments, sign-in, anything that counts. Not around everything — around the expensive parts.
Shipping an update becomes a button
One click, with the same result every time, and a way back that's just as simple.
You get a copy to try things in
The same as the live product, with made-up data. There, you and I can be wrong without any customer noticing.
Then we change what you wanted to change
The things that sat untouched for six months because nobody dared. Now a mistake costs a rollback rather than a weekend.
8 min read
How to identify potential leaks, rotate environment variables in the right order and look for unauthorised use without unnecessary disruption.
Read the article9 min read
Stripe's real-time sync simplifies reporting and troubleshooting, but product access still requires clear rules for delays and responsibilities.
Read the article8 min read
Older clients may lose access to Supabase when outdated TLS settings are removed from projects’ HTTP APIs.
Read the articleWhat waiting costs
Every change you put off is something your customers are waiting for, and what they don't get from you they look for elsewhere.
Free
Send a link to your app or prototype and what you're wondering about. You get a written review back: what holds up, what doesn't, and what I would do first.
An answer within two working days. It costs nothing and commits you to nothing.
From a client
Daniel Hjortvall
★★★★★
“I have had the privilege of working with Ketryon for several years, and my experience has been consistently excellent. … Ketryon combines strong expertise with a solution-oriented approach, which has greatly contributed to the success of our projects.”Price
The safety net is usually the first request, and it pays for itself on the first change that didn't turn into an incident.
Six situations we hear about often. Pick the one that sounds most like you.
Three questions about what you want to build and when. Then we show you what actually applies to you.