snehesht

Congrats on launching, I’m working on a similar idea with local models.

I'm working on something similar with userspace wireguard, will share it soon.

You're right but for simple use cases its useful. Someone pointed to ds4 + qwen3.8 with Q4_K will try that out.

This is interesting, thanks. - https://github.com/Neroued/ninfer

Yeah, I was surprised at first then had to dig through setup.py and setup.sh files to figure out.

They have some community benchmarks published https://github.com/Niko1221/Strata/tree/main/bench/results

Yeah I agree, I'm running it with Pi didn't notice much difference compared to lower tier models and the speed, of course.

I tried it and it worked surprisingly well. On my machine (Nvidia 4090, 128GB DDR5, Ryzen 7950x3d) I'm getting 124 tokens per sec, thought to share it here.

https://huggingface.co/Qwen/Qwen3.8-Flash-Next

I have been using jcode for past two weeks, honestly I feel its much better compared to Opencode.

This looks neat, congrats.

For anyone looking for a more general renderer, https://github.com/charmbracelet/glow.

Thank you.

What's the name of the keyboard in the picture ?

Nice idea. Good luck.

Oh my Pi - batteries included pi harness

To be precise,

"git.addAICoAuthor": "off"

Interesting idea, why not run this custom agent as part of PR Review process ?

Good clean PRs accepted based on score from your customer agent, rest are closed with comments.

If this is for AI agents to push code, I’d recommend introducing a feedback look from your review agent back to your coding agent.

True, most of the blacklists systems today aren’t realtime like Amazon WAF or Cloudflare.

We need a Crawler blacklist that can in realtime stream list deltas to centralized list and local dbs can pull changes.

Verified domains can push suspected bot ips, where this engine would run heuristics to see if there is a patters across data sources and issue a temporary block with exponential TTL.

There are many problems to solve here, but as any OSS it will evolve over time if there is enough interest in it.

Costs of running this system will be huge though and corp sponsors may not work but individual sponsors may be incentivized as it’s helps them reduce bandwidth, compute costs related to bot traffic.

You wouldn’t permanently block them, it’s more like a rolling window.

You can use security challenges as a mechanism to identify false positives.

Sure bots can get tons of proxies for cheap, doesn’t mean you can’t block them similar to how SSH Honeypots or Spamhaus SBL work albeit temporarily.

Kokuyo Stapleless Stapler

In US, you can find it on Amazon for 15-20 bucks.

Why not simply blacklist or rate limit those bot IP’s ?

50/200 Gb free plus $0.5 / Gb out egress data seems expensive when scaling out.

Wow, the whole thing (website, github repo) is down.

Make sense, but doesn't necessarily have to be an llm, just a regular dom parser will be able to tell whether an element is visible or hidden.

After the first five minutes of using it on Ubuntu, it crashed with error saying I don't have enough free memory, quick look into system stats proved that wasn't the case.

Anyway, not a great first impression. I guess I'll try again in a few months.

Isn’t this easy for LLMs to avoid by passing an instruction to ignore any hidden links ?

Thanks for pointing it out.