ProDrifterDK

[ProDrifterDK @ProDrifterDK]

@rod_herrera Pero si la esposa de Kast tiene una fundación que los regulariza. Era obvio que iba a pasar esto

[ProDrifterDK @ProDrifterDK]

@sudoPhoeniX @NousResearch @Teknium Whats the point? You already have telegram, WhatsApp, discord or any mobile app on your phone. Ergo, you can use your hermes agent on your phone

[illuminatibot @iluminatibot]

Can't do it, sorry.

[ProDrifterDK @ProDrifterDK]

@alphabatcher Skill issue and batshit

[ProDrifterDK @ProDrifterDK]

I first started vibecoding with chatgpt copying and pasting like 3 years ago. Then I started using Roo code on vs @code then I met @claudeai code and switched to @opencode with GLM and then with @OpenAI GPT 5.5. I heard from Hermes from here and there and decided to give it a try. Once I tried it, I never changed back to opencode. I really liked opencode, but @NousResearch Hermes is simply really good. It does feels like an autonomous agent and not a simple AI coding tool.
I think the ability of creating and auditing its own skill set to reuse them because he found a gotcha or a repetitive workflow, is simply amazing.
Also it's super simple to set up on almost any messaging app (telegram and discord for me).
So yeah, once I tried Hermes I just couldn't go back and it's also the main reason I can't go back to any Claude subscription since they only work on CC (not to mention that GPT 5.5 is the overall best model yet).

I just would like for the team to be able to review more PRs, but since there are more than 5K open, I absolutely understand.

Best agent ever (until another one does it better which is really unlikely, or if Nous Research really fucks it up lol)

[ProDrifterDK @ProDrifterDK]

@dom_lucre Which group of people does the only censored MJ song They don't care about us mention?

[ProDrifterDK @ProDrifterDK]

@dom_lucre Which group of people was MJ talking against on interviews?

[Dom Lucre | Breaker of Narratives @dom_lucre]

🔥🚨BREAKING: Netflix has been caught editing footage of Michael Jackson to make him appear guilty in their new documentary on his trial named “The Verdict.”

[ProDrifterDK @ProDrifterDK]

@polydao

[ProDrifterDK @ProDrifterDK]

@RogerThatJim @cyrilXBT Yup, the main UI/UX issue it has right now. I went back to the terminal after that, hoping they fix it anytime soon

[ProDrifterDK @ProDrifterDK]

@Teknium @johancutych @NousResearch Can confirm. I have 2 custom plugins (built by me) on my Hermes and after updating several times and hundreds of commits, neither has been broken by any update

[ProDrifterDK @ProDrifterDK]

@Teknium I didn't have any error while updating c:
→ Checking configuration for new options...
✓ Configuration is up to date

✓ Update complete!
→ hermes-gateway: draining (up to 75s)...

✓ Restarted hermes-gateway

[Resyst Labs @ResystLabs]

Updated LLM benchmark table:

9 models · Full/Agentic + SWE MVP.
DeepSeek leads at 91.375. Qwen/MiniMax/Claude are within 0.12. Nemotron enters with 95.29 capability but 58.63 SWE.

@deepseek_ai @Alibaba_Qwen @MiniMax_AI @AnthropicAI @GoogleAI @nvidia @NVIDIAAI @OpenRouter

[ProDrifterDK @ProDrifterDK]

@Sentdex There are some results
https://t.co/HKYgs1ssNu

[Resyst Labs @ResystLabs]

We benchmarked @nvidia Nemotron 3 Ultra 550B-A55B on @OpenRouter.

Surprise: strong general reasoning, but SWE patch production broke hard.

Full: 79.54 · SWE: 58.63 · Overall: 69.08

Great analysis model. Weak autonomous coding agent.

@NVIDIAAI #Nemotron #LLMBenchmark

[Resyst Labs @ResystLabs]

All-in-one dev platforms were built for retention, not output. Cursor hit $2B ARR by doing one thing well. Linear eats Jira by having fewer features. The unbundling isn't a trend — it's a correction.

https://t.co/DOEkoEH3YO

[ProDrifterDK @ProDrifterDK]

@atomic_chat_hq It just seems to be a little bit better on the frontend and nothing else.
It's cheaper, yes, but DeepSeek v4 is even cheaper. We'll need further testing.

[ProDrifterDK @ProDrifterDK]

@kmdrfx Bro I was really invested in knowing the latest features of opencode lmao

[ProDrifterDK @ProDrifterDK]

@cyber_razz They're asking that because they know the other AI labs are doing better