Sabitlenmiş Tweet
Tom Mulkins
757 posts
Tom Mulkins
@tom_mulkins
Janitorial Business Owner in Tampa. Pushed electrons for a while, now atoms (and electrons). Give good vibes, get good vibes 😎
Tampa, FL 🐊 Katılım Eylül 2013
335 Takip Edilen326 Takipçiler
Grok Build is also a really great interface for running subagents, which it often uses to check its own work.
English
@TheJusticeDept I hope when you arrested him he was wearing that stupid-a** outfit!
English
🚨Owner of Health Care Software Company Convicted of 1 BILLION DOLLAR Medicare Fraud Conspiracy
“The Department of Justice crushed one of the most egregious fraud schemes in Florida history,” said Acting Attorney General Todd Blanche. “This illegitimate operation stole more than $1 billion from American taxpayers — including hundreds of thousands of Medicare beneficiaries. This was cold, calculated, industrial-scale theft targeting the sick and elderly, coercing vulnerable people into buying unnecessary medical equipment. We will not rest until every fraudster ripping off the American people is held accountable.”
Read more: justice.gov/opa/pr/owner-h…
English
Just realized I ran all of this with sonnet 4.6
Wes Bos@wesbos
Burned $91.34 with Claude Code /goal in 3.5 hours Unreal, It was able to reverse engineer it!
English
@zanehengsperger lol - see if you can find someone with a Costar login.
English
request for startup: a commercial real estate broker where the building listed online is still available AND has more than 1 picture
English
Tom Mulkins retweetledi
Please everyone use Socket Firewall and set your package managers minimum release age. All node package managers as far as I found support this; set it to 7 days. This would mitigate most of your risks
Socket@SocketSecurity
🚨 UPDATE: Mini Shai-Hulud has crossed from @npmjs into @pypi and is still spreading. Newly confirmed compromised artifacts: @opensearch-project/opensearch: 3.5.3, 3.6.2, 3.7.0, 3.8.0 (1.3M weekly downloads) mistralai: 2.4.6 on PyPI guardrails-ai: 0.10.1 on PyPI additional @squawk/* packages on npm guardrails-ai 0.10.1 executes malicious code on import. On Linux, it downloads git-tanstack[.]com/transformers.pyz, writes it to /tmp/transformers.pyz, and runs it with python3 without integrity verification. The git-tanstack.com domain displayed a message signed “With Love TeamPCP,” along with: “We've been online over 2 hours now stealing creds Regardless I just came to say hello :^)” The page also linked to a YouTube video and you can probably guess which one.
English
Tom Mulkins retweetledi
Hermes Jam Session #2
@NousResearch @Teknium
feat. demos from the community:
- YANVI physical robot
- DroidRunner
- Superslick GUI
- Television/Telepath
- Cronalytics
& others that I forgot to mention!
Energy and vibes on another level.
English
@NousResearch Nous Research discord really is 🔥
many talented people in there
English
Thanks to everyone who showcased their projects in today's Hermes Agent Jam!
English
@pnpmjs be great if the community could rally around practices like github.com/lirantal/npm-s…
English
Using VS Code to type some code like a monkey. It's basically useless now? Auto-complete doens't work at all or takes forever.
English
@__RobertG__ @Teknium I had my agent setup Honcho for itself locally using containers. Just prompted it to do it then had it test the container health for a few days to make sure all was good. Been solid for a month so far.
English
Can someone who is a seasoned Hermes Agent user recommend a good memory setup? There are a number of options (Honcho, OpenViking, Mnemosyne, etc.) and I have no idea which one to choose. I just want something that works well and that I can run locally. @Teknium, any advice?
English
@nikitabier If you've ever been a business owner then you'd know this is always how it works.
English
> Took 3 days off to go to best friend’s wedding
> App is on fire
Lesson learned: Never have friends
English
@TeksEdge Can you share what top p/k, thinking enabled, and context window size?
On M2Max w/96gb I have to adjust repetition penalty to prevent thinking loops on Gemma4-35b and still not reliable yet.
English
🔥 Qwen3.6-35B-A3B is INSANELY strong 🔥
✅ 100% ts-bench success rate
(with opencode, vibe-local, GitHub Copilot, qwencode & Claude Code)
⚡ Matches Claude Sonnet 4.6 & Opus 4.6 task speed
⚡ 3x faster inference than Qwen3.5-27B → way shorter completion times
💻 Runs on consumer gear:
Mac (32GB+ RAM) or RTX 3090/4090/5090
金のニワトリ@gosrum
Qwen3.6-35B-A3Bが強すぎる！！！ ・opencode,vibe-local,GitHub Copilot,qwencode,claude codeと組み合わせたときのts-benchを実施したところ、すべて満点 ・しかもClaude sonnet 4.6やOpus 4.6と同じくらい速くタスクを遂行できている Qwen3.5-27Bもすごかったが、Qwen3.6-35B-A3Bは赤い彗星のごとく27Bよりも推論速度が3倍速いので、ベンチマーク結果からもわかるようにタスク遂行までの時間が大幅に短縮できるようになったのが大きい
English
@ThePrimeagen Draconian linting (use lintstaged), type checks, tests, then write up the coding agent with your workflow.
Review agent checks it after coding agent says "done" like "oh really, let's make sure you did." Hands back results to coding agent.
English
I am slowly coming around to AI assisted programming.
I am genuinely trying to codify every rule about programming that I have and using that + several stages to build out small changes.
Not sure the productivity changes, but I think I can see a modest gain in speed. I am also trying to be concerned about every line produced, not just slop trebucheting code over the wall.
English
@adilmania Is this real? Because I'm watching the shit out of this if it's real
English
Tech has a narrative problem.
its either pure hate (old + social media)
or pure in-the-bubble (new media)
i'm fighting to change this.
cause i believe the future will be bright ☀️
Tech needs its own Night Show. something simple, fun, entertaining, relevant. so anyone can actually understand what's really happening in the best way possible.
Sebastian Thrun (Waymo, Udacity) was first.
(only in-person due to some technical issues, all fixed for the next ones)
Who do you wanna see on the show next?
Adil Mania.@adilmania
introducing Silicon Mania Night Show. the first tech late-night show, live every Saturday on X, YouTube & Twitch. tech got boring. time to make it fun again :)
English
@pmarca Smaller and older models could find them too.
"...on a basic security reasoning task, small open models outperformed most frontier models from every major lab."
aisle.com/blog/ai-cybers…
English
Every single security hole discovered by AI was already there.
English
English
We have partnered with @Xiaomi to bring their excellent MiMo V2 Pro model to Hermes Agent via the Nous Portal - completely free to use for the next 2 weeks!
Access now on the latest version of Hermes Agent: 'hermes update'
English
@sandipb @MozDevNet Yeah, I see from the example picture they didn't use the syntax from the MDN doc:
```
parentRule {
/* parent rule style properties */
& childRule {
/* child rule style properties */
}
}
```
English
@tom_mulkins @MozDevNet But that requires no white space between & and child selector, I think?
English
🧩 Write cleaner CSS with nesting
You can now nest selectors directly in CSS.
It improves readability, keeps styles modular, and reduces repetition.
No preprocessor needed.
Learn more 👇
developer.mozilla.org/en-US/docs/Web…
English