I stripped my credentials out of the file, but here is what I made using claude opus5.5 and my Ventuno Q:
This was all built ontop of the offline llm example, it starts with a splash page, then you can select other programs from the web UI, you can set effort levels in chat, have it use a cam in chat to tell you what it sees, you can give it personal info to get a better suited response tailored to how you learn, on the ventuno you can run vision and chat at the same time. You can also monitor real time logs and board sensors. It got buggy after yesterday's update, I haven't tried this on uno q (which I have), this was built on the Ventuno q
Splash
Monitor page: pin is 1234 default
Vision
Chat With selectable effort levels, and settings like ui appearance, and user info, which crafts a backend system pro.pt to make answers suited to how you read and learn
Quick settings to also use a network cameras or USB for vision:
Section to help customize answers for you:
On the stock file I removed my credentials, and on first run go to boards Ip :7001 port to setup how tou want tou menus and options, then main page is accessible at yourip:7000
This all runs off one app lab program, and has been working well. On an uno q i would imagine it should work but be slight slower (have not tried it on my unoq yet)
Also: here's a list of what I had to debug after yesterday's board and applab update:
Stock file to import to app lab, run on port 7001 first use
llm and vision multi program.zip (628.3 KB)
Lmk how it works!
-aJ
