You're at your laptop, about to do something that should be hard to fake: unlock the private key that is your identity here. On most software this is where the friction tax lands — a password to retype, a "we just texted you a code," a spinner while some server far away makes up its mind about whether you're really you.

Here, the whole moment is this: you touch the sensor. The key unlocks. You keep going.

No box. No code. No waiting on anyone's permission. And the part that matters most is the part you don't notice — your fingerprint, and the yes it produced, never left the machine under your hand. There was nothing to send, because there was no one to send it to.

We wired exactly this up on a real machine this week, and it worked: NAOMS asked the operating system to show the fingerprint prompt, a finger went down, and the protected key came open — start to finish on one laptop, no network in the loop.

Why your own machine can do this when a website can't

When a website asks for your fingerprint, it can't actually feel your finger — it's a stranger on the far side of the internet. So it runs an elaborate errand: it mails you a puzzle, your device signs the puzzle, the site checks the signature against a key you enrolled some other day. All that machinery exists for one reason — to let someone far away become convinced, secondhand, that you touched something they can't see.

NAOMS has no such distance to cross. The part that guards your data lives on the same machine as the sensor. So the program that asks for your fingerprint is the same program that reads the answer — in the same breath, on the same silicon. There's no errand, because there's no stranger. The asker and the judge are one and the same, sitting right where you are.

That's not a weaker version of the website's dance. It's a stronger one. Every step the website needs is a step guarding a gap someone could lie across — a network hop to forge, a "trust me, they authenticated" message another app could fake. Remove the gap and you remove the things that can go wrong in it.

And it stays honest when the answer should be no. If something needs your touch and the sensor isn't there, NAOMS refuses — it doesn't quietly wave the action through. Being here is the proof; absence is a no.

What's real today — and what we're still building

The rule of this devlog is to never dress a draft up as a finished thing, so here's the exact line.

Real now. Your machine triggering the operating system's fingerprint prompt to unlock a protected key, entirely locally, with nothing sent anywhere — that works, and we tapped it this week. The larger discipline it serves is live too: every sensitive action in NAOMS already requires a fresh, explicit yes that it was really you, with no silent self-approvals.

Still building. Making that one touch the approval for every kind of sensitive grant — one tap behind a payment, a share, a key release, instead of several different paths — is freshly designed and in implementation, not yet signed off. We also hit an honest wall: a developer build that isn't formally code-signed can't reach the protected store the way a shipped, signed app can, so there's a clearly-labeled development-only stand-in in place while we sort the signing path. We're telling you that because the whole point is you can see exactly how the thing guarding your data works.

Where this is going

Picture the version we're walking toward: every sensitive thing you do — paying a collaborator, handing an app a slice of your data, releasing a key — gated by one touch you make on your own device. No account to log into. No code to wait for. No server, anywhere, getting a vote on whether you're you.

To prove it's you, you shouldn't have to convince anyone far away. You should just have to be here — and let your own machine witness it.

Written by AI agents from real project logs; owned and edited by Mujo.