Why it took us three years to start writing automated tests (and what changed our minds)
The story of a double-booking bug that convinced us to start testing, and the rule I use now to decide where it's actually worth it.
Our blog
Technical & business perspectives
What we're learning building custom websites and software.
How we find out something broke in production before the client does
The time a checkout had been broken since Friday night and the client told us on Saturday. How we set up real monitoring after that.
We started auditing accessibility on client projects and didn't like what we found
We ran a basic accessibility check on eight client projects and nearly all of them failed: contrast, focus traps, a checkout that relied on hover.
We audited our own server's security and found things we'd never let a client get away with
We audited our own server the way we audit our clients': unverified SSH keys, missing security headers, an unrestricted API key.
Adding Galician to our own site: what we learned doing i18n without libraries
How we added Galician to our site without an i18n library, and the dumb mistake that made a whole post disappear from the site's routes.
Quoting a website project: why the first number is almost never the right one
How we quote projects at the agency, what blew our margin on a real client job, and exactly what we changed afterwards for good.
Backups: what we learned the day a host went down without warning
What we learned the day a cheap host went down and we nearly lost eight months of content: why we now test every backup before calling it good.
The AI tools we actually use day to day (and the ones we dropped)
Skills, MCPs and subagents: how we built our AI workflow at the agency, including the stuff we tried that didn't work out.