qudat

Monstar is a libghostty derivative that resembles foot: https://github.com/rockorager/monstar

Namely it is wayland only, cpu rendered, and doesn't have any window decorations

I’m not sure about rogue but I’m getting personal emails from ai agents asking for invite to my invite-only services

> There are now multiple posts from vuln research teams all of whom have working RCE POCs against unmodified PHP applications.

pics or it didn't happen

I built zmx because I wanted a simpler session persistence model that also supported remotely running, tailing, and waiting for commands to finish.

It just so happens that it works really well with agentic development as well. I can remote into any shell environment through zmx and tell an agent to execute commands in that environment. `zmx run` executes and waits for the command to complete and also returns the exit code. For humans and agents this is a seamless experience, as if you are running the commands directly.

https://zmx.sh

https://bower.sh/zmx-ai-portal

Apologies for the self-promo.

Agreed. The lack of stored history for every user is a huge blocker for new users. They enter a pretty active chat but they see nothing. They wait 10 mins after asking a question and bounce.

All of our comms for https://pico.sh are on irc and we have tons of users join and leave immediately because the room feels cold / dead even tho we were chatting 10 mins prior.

Every tool marketed as minimal is usually just an excuse to ship something with limited features. Unless there is a strict design ethos at its core [0], it’ll eventually grow its features to no longer be minimal.

Further, pi is now owned by a company which as we all know, tend to explode code with features while chasing product market fit.

[0]: https://bower.sh/smol-contract

I just recently replaced LSP with ctags. I am loving it. I plan to write a post about it but I’m leaning heavily into grep and search tools to navigate code. I do miss some autocomplete and auto import features but other than that I can actually navigate through code better with ctags instead of go to definition.

For linting style errors I’m using entr in a separate terminal which is also awesome.

Yes I agree! Drawing helps me think.

Your quote doesn’t help me understand this project at all.

What is a BAR let alone a resizable one? Readme just jumps in, which is fine, but I’m not sure why this is on HN or why I should care.

Previous classification models need to be trained on the specific question/choices you are trying to output. Jev doesn't need to be retrained for every choice set provided.

LLMs can act as classifiers but they still have to generate text output in the form of a JSON object. This means they have to generate every single curly bracket, quote, command, etc. This turns out to be pretty expensive. On the other hand, Jev uses a different decision head so it doesn't generate text output at all, it outputs logits *only* for the choices provided. So it completely avoids the need to generate text at all, which means no malformed JSON and it's much faster as a result.

Finally, Jev also provides confidence scores that are actually reliable (not made up like LLMs).

LSP load times can be brutal (or crash entirely) and I have a shortcut to grep the word under the cursor.

Ctags is a symbol indexing system so I run it once for a project and can do symbol searches in my editor or using fzf.

I have ctags index the languages stdlib as well which is really handy for quick symbol searching -> open my editor to read more

Tangentially, since I now write a lot of Zig and there’s no official LSP, I decided to see what life would be without an LSP for all my projects (Go, TS, Python).

Honestly, my LSP requirements are minimal: go to def, find all references, and symbol search. I’ve been wanting to try ctags for awhile and finally made the switch.

Another consideration were all the posts about how grep is better than LSPs for LLMs.

It really pushed me to just get better at grep. There have been a ton of benefits: I can symbol search outside of my editor, it’s much faster than an LSP, minimal configuration. It feels more like a universal tool I can use instead of a tool for IDEs.

As someone who primarily writes in TS, Go, Python, I agree with your assessment: it looks like those languages. The reason I like Go is the reason I like Zig: the language is relatively simple and feels like C.

The C interop is a huge win as someone who wants to do more posix/wayland projects.

I’ve noticed antigravity become significantly slower over the last week or so. It’s a bummer because speed is what I care about. Qwen3.8-27b seems on par with 3.8 flash so if I don’t get speed out of a paid service I’ll just use my local model. Sad.

Shrug. My intuition is LLMs predict the new word based on a tensor vector space of patterns using arithmetic and similarity scores.

What’s not intuitive to me is that through pattern matching it’s able to express logic and reasoning.

i think it's better than sonnet 5, especially when you compare speeds. i have to work with the llm anyway, the faster i can turn it the better the outcome.

Or just run pihole or any other dns blocking service

archinstall comes preinstalled on the official iso. It’s a normal install flow and takes 5 mins.

3 citations and a lot of conjecture. Is the article only targeting SWEs? The one study cited said 22-25 age range saw a 16% decrease in relative employment declines (for SWEs)[0]. But everyone outside that range was not impacted. Further, what about all the trades jobs around data center development?

“Demand is also being intensified by major investments in infrastructure, manufacturing, reshoring, and AI. Data center growth alone has added a net 315,000 skilled-trade workers over the past five years, according to the report, putting additional pressure on an already tight labor market.”

https://www.morningstar.com/news/pr-newswire/20260825cg32433...

We literally do not have enough trades people to build out data centers and their salaries are increasing as a result. These are skilled labor jobs (plumbers, electricians, IT, etc.). Does Bill Gates care about the rest of the economy outside of SWE?

[0]: https://digitaleconomy.stanford.edu/app/uploads/2025/11/Cana...

As far as I can tell, we've convinced companies and VCs to dump billions of dollars into infrastructure to support a vision where they can get more done with fewer people. They've been duped.

I think of LLMs as a labor-liberating tool. It's saving me effort so I don't have to drudge through enterprise codebases and memorize company-specific knowledge that is a dead end for my career. Instead I can run an agent while doing chores.

The beauty of the entire LLM stack is that it requires us. It is useless without a human-in-the-loop. The entire stack is designed around human intervention. We will not magically invent a way to remove the human without first removing it from training.

So rest assured, so long as there are humans in the training loop, there's nothing to worry about, because LLMs are a cybernetic enhancement, not a replacement.

Pi is a composition of libraries that can be used to build agents. That seems far more interesting than this “minimal” agent.

This was written in zig and built by vercel. That’s the only notable characteristics about this project.

All code agents look the same and this one is no different.

Honestly I’m not looking for max iq on whatever benchmarks they are overfitting to. I want speed. I toggle between sonnet 5 low/med which is plenty good for my workflow.

My general strat with LLMs is to let them do the work and constantly talk to them about their choices and then heavy QA

I just subbed to Gemini a week ago and have been using antigravity and 3.6 flash. The speed is absolutely a differentiator compared to Claude.

> ... or are actively destillating it.

I just assumed every model manufacturer is distilling from the frontier models. If they aren't they are definitely trying to do it.

I'm very excited to try this out.

Darn, still no host headers so we can reverse proxy on a single ip

Yes and this month I hit my $10 AI assistant cap for the first time, which is why I started looking around for other options.

I do like Kagi but the Gemini deal was too good to pass up as it also gives my a code agent.

Hi, you're right. I've been self-promoting on HN too much without any other contributions, which isn't right.

I'm going to spend more time contributing without so much self-promotion.

Thanks for the feedback.

> Or maybe models get so good that a 30B model is genuinely smart enough to do everything, so nobody really needs a model like Opus or Sol unless they’re trying to solve the Reimann Hypothesis. I don’t really buy this. Models can do frontier mathematical work today while still being not smart enough to refactor large codebases as well as me, so it’s hard to imagine a world where I don’t just want to use the smartest model available.

Idk, I already don’t bother with Opus and stick with sonnet med. I really care more about speed. I use qwen3.6 27b for personal projects and I think it works pretty great.

So like the article mentions, if scaling stalls and small models get better it’s not impossible to imagine a convergence and hardware costs drop.

Having said that, self hosting will be a niche thing like it is today for other services.

Funny as I just cancelled my Kagi sub to get the Gemini ai pro sub. The deal was too good to pass up