LouisvilleGeek

Thanks for the heads up!

Not sure of why everyone is negative against this tool?

I quite like it. Thanks for sharing!

I have been testing it on my security / firewall appliance (nfSensei) and it's much improved. It's actually proceeding through a complex plan Opus 4.8 proposed.

But then again I have not tested it without a prior model planning as of yet.

Note: fable would fail immediately on any security related topic prior.

Can we get Counter Strike? :)

Same here. It's been rushed for the IPO (in my opinion).

Call it Ziggy

Finally a use for the expensive ram when it's not needed in workloads!

Now if it could be dynamically used and vacated on other GPU workloads?

Amiga 500 + Fidonet brings back such fond memories.

Might as well stay off the internet on April 1st.

Can confirm. You need to kick off the sessions to read the ~/.claude/CLAUDE.md file. Annoying.

And also the recent /compact bug.. is anyone else seeing that you cannot compact and must /clear ?

Curious, how did ClawdBot get in touch with another ClawdBot that happened to live two doors down?

You should check out the new beta installer!

https://www.youtube.com/watch?v=vbWOLzGDa4I

It worked perfectly on a 12+ year old DELL laptop would think it would work for your desktop as well!

Self host with Headscale.

Looks "Glamourous" but lacks the basics:

Up / down history Copy text

Other than these issues feels much nicer than Claude Code since the screen does not shake violently.

20% CPU use with absolutely no "vt" defined commands is not exactly something to celebrate!

Also when using this tool on a iPad its completely flooded! You need some kind of diff tool when painting the screen!

But either way celebrate on your great "vibe" coding working but poor code!!

* edit! CPU not GPU. idle this thing is a pig. and when used, look out you cannot even change tabs when using claude code. How ironic. (sorry for my random train of thought)

Sorry to be blunt, vibe coders need to spend more time TESTING!

The barrier to use this project is maddening. I went through all of the setup instructions and getting the workspace error for a personal gmail account.

Googlers, we should not have to do all of this setup and prep work for a single account. Enterprise I get, but for a single user? This is insufferable.

Bambu Companion (unfortunately iPhone only) does more than Handy and works on LAN. Using it today in fact! https://www.youtube.com/watch?v=LZpDQN9zgUI

I would absolutely love to take this for a spin. Using Fusion 360 quite a bit lately.

Same here. Please consider a higher quality option.

Absolutely mind blowing.

Coming soon! https://migswitch.store/products/mig-switch-dumper-pre-order

This GPT was pretty handy for generating OpenAPI specs from docs. Used it to generate specs for Microsoft Graph and it just worked on the first attempt pointing at microsoft documentation.

https://chat.openai.com/g/g-TYEliDU6A-actionsgpt

Tried this out last night and kept getting a key is not authorized error even though the key absolutely exists. (I had just created it via the steps on the Github page)

It's unfortunate we lost Google in Louisville. Would love to have this type of offering.

That would explain a lot! Sometimes when it's fast it spits out all code. When it's slower, it's Lazy! Thanks for the link

Can you share a link to that x/tweet?

Unfortunately it's not building per the github instructions.

warning: `rustgpt` (bin "rustgpt") generated 15 warnings error: could not compile `rustgpt` (bin "rustgpt") due to 19 previous errors; 15 warnings emitted

Using actions is painfully slow this morning. Responses are slower than a 300 baud modem.

One of the things I did not care for is setting my plugin hints from the Custom Instructions setting. It seems this will allow for each "GPT" agent to have their own system instructions which will be quite handy.

Curious if anyone that has access can verify that we can still build "actions" via localhost like we previously did with plugins?