Why ContextPass.

Most products will ask you to trust them with your data and pretend they're using it for your benefit. I started ContextPass because I'm tired of that being the default.

The thing that broke it for me.

I was buying a new pair of shoes online. Same site I used six months ago. The site had no idea what size I was. So I typed it in. Again.

Meanwhile, three ad networks had already loaded eight cookies, two tracking pixels, and a fingerprinting script. They knew the last fifteen sites I'd visited. They knew the route I drove that morning. They didn't know my shoe size.

That's the model the open web has settled on. Track what they shouldn't. Miss what they should. By design, not by accident.

The pitch I keep making.

Personalization without profiling.

A wallet you own. Sites have to ask, by name, for the specific fields they want. You approve, get a receipt, and revoke whenever. They never get "the whole person." They get exactly what's needed for the thing you're trying to do — your size, your budget, your style — for one purpose, for a short window, for that one site only.

Each merchant gets a unique pairwise ID. They can't link your activity to other ContextPass-enabled stores. There's no global profile. There's no data broker pipeline. There's no surveillance on the back end pretending to be personalization on the front end.

Why the timing matters.

AI agents are about to do a lot of online shopping on our behalf. They'll need consent. They'll need scoped tokens. They'll need an audit log. They'll need someone to build the consent layer for the agentic web.

I'd rather it be a user-owned protocol than another data broker.

The window for this is narrow. The shape of the next ten years of e-commerce gets decided in the next eighteen months. Either we ship a privacy-first consent layer that becomes the default, or we wake up to a world where AI agents profile us harder than human marketers ever could.

How I'm building it.

Solo. In public. The repo is private right now but the thinking is public on this site. I'm starting with sneaker and apparel Shopify stores because the personalization signal is dense (size, color, style, budget) and the trust gap is large (every shopper has been creeped out by ad retargeting at least once).

The wallet is local-first by design. Your preferences live in your browser, not on my server. Even I can't see them. That's not a limitation — it's the entire point. Cross-device sync is a Phase 3 problem. For now, the wallet works on one device, and that's fine.

The points system rewards usage without surveillance. Points stay in your wallet, never sent to merchants or to me. Loyalty without profiling, just like the rest of the product.

The ask.

If you're a user who wants to try the wallet: open contextpass.io/wallet. Set up your preferences. Walk through the demo. Tell me what's broken.

If you're a merchant with a real catalog and shoppers who'd rather not be tracked: I want to talk. Free pilot slots are open. The Shopify app is on the roadmap. Email me directly.

If you're a builder who cares about the shape of the agentic web: come build this with me, or build something adjacent. The protocol is going to be open. The consent layer is too important for one company to own. I'd rather convince three other founders to build the same primitives than try to capture the whole market.

Either way — thanks for reading. The product is real. The site is real. The signups go to a real database. Drop your email if you want to know when it's ready for you.

— Jeffrey Thompson
Founder, ContextPass