Three weeks ago I shipped something with an error in it.

Not a catastrophe — the kind of error that would have caused a week of confusion for a handful of people, a few annoyed messages, and a sentence in a meeting that starts with "why did this happen." I caught it myself, quietly, before anyone else could. I fixed it in about forty minutes. The system is fine. Nobody knows.

And I have thought about it every single day since.

Here's the detail that tells you something is wrong: I check the thing every morning. Not to verify the fix — the fix works. I check it to see whether anyone has noticed that it was ever broken. I've reread the same thread looking for a question that implies discovery, the way you reread a text to see if they're mad.

Nobody noticed. Nobody is going to notice. That's not the problem. That is the problem.

The three reasons I gave myself

What the reading said

I did a reading. The question: should I admit a mistake nobody noticed?

The reading didn't argue about consequences. It asked: who are you protecting?

First answer: the team. The project. Everyone's week. All true and all available in a sentence.

Second answer, once I stopped performing for myself: me. I was protecting my reputation as the person who doesn't make that kind of mistake. And that was the thing worth looking at, because I didn't choose that reputation — I inherited it. Somewhere along the way I became the reliable one: the one who catches things, who checks twice, who other people hand things to. Being that person is genuinely most of my professional identity.

Which means admitting this isn't admitting an error. It's admitting I'm two people: the one who broke it, and the one who hid it. The second one is the one I couldn't stand to introduce to anyone.

And there's the arithmetic I'd been avoiding, which the reading put in front of me plainly: if this ever comes out later, the story is not "he made a mistake." The story is "he made a mistake and sat on it for weeks." The cost of confessing grows every day I don't — which is exactly why reason three feels so reasonable. It isn't a reason. It's a bet.

What happened

I told them. At the end of a standup, no drama, four sentences: here's what broke, here's when I caught it, here's what I did, here's the process change that means it can't happen quietly next time.

The room was fine. My manager asked one follow-up question — about the process, not about me. Someone said "good catch" and moved on to the next item. The whole thing took forty seconds.

What I'd been carrying for three weeks weighed forty seconds.

The interesting part came after, in a one-on-one, when my manager said something I've thought about since: "You didn't have to tell me, and I want you to know that's exactly why I trust you with more." Nothing about my standing changed. Something about my standing solidified, and it turned out the way to solidify it wasn't being flawless — it was being the person who says the difficult thing while it's still difficult.

Anyway. The fix was never the hard part. The hard part was finding out that being trusted isn't the same as being flawless — and that the only part of the trust that's actually mine is the telling.

If you're sitting on something nobody knows — a mistake, a shortcut, a number that's slightly wrong — and the only argument for silence is that it's over, try a reading. Three numbers. No account. It won't confess for you. But it might show you who you're actually protecting, and whether that person needs protecting.

FAQ

Q: Should I tell my manager about a mistake no one noticed? A: If the fix is done and the risk is gone, you're not deciding whether to report a problem — you're deciding whether to be a person who hides things. The second one compounds and the first one doesn't. Say it plainly, in one short message, with what you learned, and don't perform the guilt: long apologies make a small thing into a big one. See also: Should I Tell My Friend Their Partner Is Cheating? I Found Out Last Week. — same question, different room: what does it cost to be the one who says it, and what does it cost not to.

Q: What are the actual risks of confessing something no one caught? A: Realistically less than your head is promising you. In most teams the outcome is a process change and a nod. The genuine risks are worth naming honestly, though — a manager who keeps score, a culture that punishes transparency in practice even while praising it, a moment where it becomes the example in a meeting. You know your environment better than I do. But weigh it against the version where it surfaces in six months, attached to a story about concealment.

Q: How do you admit a mistake without losing credibility? A: Three moves. Be specific and short — what broke, when you caught it, what you did. Own the part that's actually yours without inflating it into a character flaw; over-apologising reads as instability, not accountability. And bring the process fix, because that's the sentence that converts a mistake into competence: here's the thing that changed so this can't happen quietly again.

Q: How does a reading help with a decision like this? A: It separates the act from the identity. I thought I was deciding whether to report an error that had already been fixed. I was actually deciding whether I could let people see that I'm the type of person who makes mistakes — because I'd built a professional identity on not being that type. The reading named it, and once it had a name, the telling took forty seconds.