One category that seems genuinely useful is agents that help reason about complex systems rather than directly execute tasks.
For example, I’ve been experimenting with a small tool that takes a description of an AI product and decomposes it into tasks, estimates token usage, and models the cost implications of different architectures.
It’s not an autonomous agent, but it’s been surprisingly useful as a “thinking tool” before building something.
If this is implemented fully in the browser, the file wouldn’t actually be uploaded anywhere. Modern browsers can read local files via the File API and process them with JS/WebAssembly.
One way to verify is to open the Network tab in DevTools and see if any requests are made when editing the PDF.
Interesting idea. In payments there’s been a lot of discussion around using FIDO/WebAuthn for phishing-resistant authentication. Turning the “user presence” check into a game mechanic is clever. Curious what motivated building it.
Nice implementation. Out of curiosity, how are you handling multiplayer state sync?
For a board game like this I’d imagine sending move events and letting clients recompute the board locally rather than syncing the whole board every turn. Curious what approach you took.