The protection I switched off to make CI pass
I sent the bouncer home so a Go test could get in. It turned out he'd been turning away malware since the summer, and he wouldn't come back.

I sent the bouncer home so a Go test could get in. It turned out he'd been turning away malware since the summer, and he wouldn't come back.

A release token showed up in use from four machines in AWS that I couldn't account for. Two hours later it turned out to be my own runner fleet, and I'd fixed the real problem anyway.

Everything on AWS here runs on credentials that last minutes. Cloudflare I'd been clicking by hand, and bringing it in meant living with a long-lived secret and making it as small as I could.

That morning I put a lock on a bucket. By teatime I'd taken it off for good, because it guarded the risk I'd already covered and locked in the ones I hadn't.

The homelab finally ran out. Not disk this time, capacity. So the overflow went to AWS, on the condition that it costs nothing when nobody's pushing.

Setting up an AWS billing alarm before you need one, with a KMS gotcha in the middle that stops the alert ever reaching you.

A fresh AWS account cannot enable GuardDuty or Security Hub without a subscription, so the security baseline failed on the account it protects.

Two CI and infrastructure bugs with one shape: a GitLab job with no rules block skips merge requests, and an import block runs only once.

Closing the plan-versus-apply gap in infrastructure CI by applying the saved plan artifact rather than quietly re-planning at apply time.

Why an infrastructure repo stays one OpenTofu graph rather than splitting into per-concern stacks: the engine enforces ordering for free.
