🤣

If Mythos is so great at code reviews and finding bugs, how come claude code and the claude website has so many bugs and shitty uptime? 🤔

Anthropic released their Mythos preview scores and its 🤯

#BuildInPublic

GitHub - milla-jovovich/mempalace: The highest-scoring AI memory system ever benchmarked. And it's free. The highest-scoring AI memory system ever benchmarked. And it's free. - milla-jovovich/mempalace

The Resident Evil actress open sourcing an AI memory system was definitely not on my 2026 bingo card but here we are 🔥🔥🔥

github.com/milla-jovovi...

GitHub - rtk-ai/rtk: CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies - rtk-ai/rtk

Checkout the github repo at github.com/rtk-ai/rtk

Been using rtk to cutdown on the amount of tokens consumed by common CLI calls for the past 3 days, and it's already saved me >6M tokens. Haven't noticed any degration or issues, and love that you basically set it once and forget about it. Have a look!

#BuildInPublic

Woah! $200 Claude credit! Check your usage tab!

It says “failed to apply” if you click on it, but you can see $200 appear on your extra usage balance at the bottom

Thanks Anthropic 🫶

#buildinpublic

AAAARRRRRRGHHHHHH!!!!!

huh....

Claude’s been hella slow the past day or so. New model incoming? Usually happens when they start taking resources away for a new model from what I gather?

a man in a suit and tie is sitting at a desk with a computer and the word and written on it ALT: a man in a suit and tie is sitting at a desk with a computer and the word and written on it

checking on my claude quota after a couple of prompts be like

So… how many of y’all panicked and checked your source map config #buildinpublic?

What’s the most important step a person can take? It’s not the first. It’s the next. It’s always the next.

Petition to change ramen profitability to AI subscription profitability

#BuildInPublic

Sorry Jonas, already gave it away earlier in this thread

But I won’t! 😂

Claude to me is definitely more fun to use, feels more humane while codex feels more robotic. Codex does catch things that Opus misses and vice versa, but both are still very very good

Sorry James, gave it away to Casper earlier in this thread! Will let you know if I get access to any more

Fine, I guess I'll go have an early night then...

DM’d it to you! Have fun!

I've got one more Claude guest pass (1-week access) to give away #buildinpublic, let me know if anyone wants it, will DM it to you

One Dark Pro theme with Vivid colors now available as a #zed extension! LMK what you think!

zed.dev/extensions/o...

😂

is this the new "it's compiling"?

How Claude feels when you say “Make no mistakes”

DM'd it to you Dom! Have fun!

Got 3 Claude guest passes to give away. Reply here if you want and will send it through #buildinpublic

Added the following to my global CLAUDE md and that seems to fix it:

**Always** use built-in tools over Bash equivalents

Added the following to my global CLAUDE md and that seems to fix it:

**Always** use built-in tools over Bash equivalents

Claude code seems to love trying to use grep, cat, find etc on the CLI instead of the inbuilt tools causing so many permission prompts during plan mode. Started happening a few days ago. Is it a known thing or something wrong on my side?