Big feeling, huh.
That's Lur — the little feeling, the blur that stays. Hlur makes small, carefully tested things for big feelings: a plain-English reset on AI, apps that calm you down, books that explain you to yourself, and stories that stay with you.
Try the smallest thing I make.
One breath in, one breath out, five times. Lur breathes with you. No app, no account, no tracking — it's all right here.
Small experiments, tested until they're boring.
AI literacy, apps, books, stories, even finance tools — different shapes from the same one-person lab. One method under all of them: agentic AI as the build partner, a semiconductor engineer's testing standard on top. Every row ships behind its own gate.
AI literacy, from one calm page new in the lab
What AI really is, in 10 peelable layers — plus a next-word demo you can press Play on, five hands-on deep-dives into how it actually works, and a 9-question check of where you stand. Model and cost facts carry the date they were verified. Built for everyone: young or old, techie or not.
Apps that calm
One-minute tools: a 30-second reset (you just tried it above), a lens for catching the mental frame you're stuck in, and a plant checkup that turns a photo into a care plan.
Books that explain
Plain-language books on the universe, the mind, and intelligence — written so anyone can read them, cited like papers.
Stories that stay
Ten original characters in short animated stories with real-life lessons.
Soft on the outside, honest on the inside — like their little bios:
Said
I'm listening.
Wasn't.
w
w
.
h
l
u
r
.
ai
Tools that tell the truth private lab
The private half of the lab — finance tools built for one user: me. You can't log in,
and that's the point: the same standard that keeps my own money-tools honest sits behind
everything else here. Eleven years of backtests, one dull verdict: doing nothing won.
The plush rides a robot vacuum.
Hlur started as a question — could one person, with agentic AI as a partner, build small real things that make a big feeling easier to hold? Everything here is an answer. AI helps build all of it, but it stays underneath, never on top. What's on top is checked the old way:
Tested until it's boring
Every app ships behind a test gate. "It looks right" never counts; only a passing run does.
Measured, not hyped
Tools grade themselves against reality. When the data says "not proven," they wait. Waiting beats pretending.
Held, not fixed
Lur never fixes anyone. The apps don't promise transformation — just 30 honest seconds, and whatever you do with them.