
Ver_Teller
7.8K posts





🇮🇱🇺🇸🇮🇷 Tehran under massive bombardment from the Epstein coalition. But Iran has hit Israel and the US/Gulf much harder today. But don't worry Karoline Leavitt fighting back tears has declared we already won.. stop hitting us back so hard Iran "Any violence beyond this point will because Iran refused to understand they've been defeated" 😂🥲


ОАЭ аннулировали ВНЖ для всех граждан Ирана, присвоив иранских активов на 530 миллиардов долларов, включая деньги, полученные за «золотые ВНЖ» на 99 лет Об этом сообщило издание Tehran Times. Представители Эмиратов пока официально не подтверждали эту информацию. Иран начал бомбить объекты ОАЭ после начала американо-израильской операции. Тегеран заявлял, что атакует американские базы на Ближнем Востоке. В четверг падение обломков сбитого беспилотника в Абу-Даби привело к гибели двух человек.


THIS IS A BIG DEAL: Congressional Investigations Confirm That the US Government, Working With the UK and Israel, Has Been Secretly Funding and Controlling Al-Qaeda, ISIS, Boko Haram, and Other Islamic Terror Groups for Decades USAID Documents Prove Senator John McCain and Lindsey Graham Have Been the Bagmen for the Global Islamic Jihad for Over 30 Years Graham’s Congressional Challenger in South Carolina Is Now Openly Calling for His Arrest for Treason WATCH LIVE: ⏬ x.com/i/broadcasts/1…







Switched from OpenClaw and honestly it's night and day. OpenClaw had the theory and the hype, but execution felt like a buggy mess despite all the resources thrown at it. constant friction, unstable tool calls on local models, too much complexity. Hermes on the other hand has been damn near faultless. Super straightforward setup, just works with Ollama (I'm running it on Ollama and letting Hermes auto-switch models based on task complexity, fast lightweight ones for simple stuff, stronger models for heavy reasoning/tool work). It actually figures things out, experiments, builds persistent skills/memory on its own, and keeps improving over time without me babysitting configs.Feels way more intelligent and useful in practice. The self-improving loop is the real winner. Highly recommend giving it a proper side-by-side on the same model/hardware.



















