Why is this easier or more effective than just a /docs directory?

You absolutely can use "just a /docs" directory in Fossil. You can even point the web server to /doc/trunk/docs/index.md or whatever other file names you want. :-)

Parent comment linked to that answer.

Interesting - so is this a local llm that identifies scams and prevents me from clicking? Seems like an awesome opportunity for a local version of Jev to say “scam/not scam”!

What are peoples' thoughts on whether a local version of Jev is possible? Having to call an API for something that's main benefit is speed is orthagonal to their ethos.

This is great! Whats the tech stack?

I started thumping cars but then got into an infinite spiral I couldn’t get out of.

three.js and Vite, very basic, I set out originally to make a game where Neil knocks over cones and grew it from there. If I was to start again I would use a game engine as it has scaled further than I intended.

What do you mean by "infinite spiral"?

This is really cool. I wish the example stories let me see the entire book and purchase them if I like them.

I’m skeptical about the stories being good quality so seeing the full stories might mitigate that.

The story synopsis next to the preview gives you the full narrative arc before you commit. But fair point on wanting to see more.

You can edit or regenerate pages if something isn't working - it's iterative, not one-shot. Happy to help you try it out without payment - drop me an email.

Not being mobile responsive made me bounce

You should just launch it no? I’m looking to try one of these and getting to market fast is important because there will be many more very soon.

The use case examples are good, would be nice to include a couple that use browser automation or that feel more magical than reminders.

Is the memory functionality just what’s baked into OpenClaw/Pi or is it customized somehow?

Yeah, it’s pretty much ready, I’m just squashing a few last bugs. Launching this week, trying to move fast for exactly that reason.

Good call on the examples, I’ll add some that show off browser automation and more complex workflows.

Memory at launch is what’s baked into OpenClaw, but I’m planning to upgrade it to vectors + a continuously updated doc shortly after (similar to what Claude Web does)

If you want early access I’d be happy to get you set up personally, just shoot me an email at ramon <at> agentmode.co

How do you create the cards for Kardy? I wonder whether just making great ecards wouldn’t be a bigger market.

Interesting - is it just the URL or is it actually crawling the phishing site to assess whether its phishing? And how does it distinguish the site bank.com from bąnk.com (with a little curly ą) if both are an identical clone of bank.com?

That's a great question - right now it is only looking at the results from a battery of several dozen indicators that we compute upstream of the model itself (which saves massively on tokens)

As small models continue to improve, and edge hardware becomes more capable, we would really like to run larger models that could incorporate full page content and screengrab data, which would be more likely to catch these kinds of attacks.

But we also find that sites that do one shady thing usually do others, which is a big reason why a tiny model like this can work - and why we are betting on low latency being a differentiating factor in real-world impacts.

Is this going to put contextual ads into my codebase or into the discussion if I use it with my coding agent of choice?