Using an "Agentic Ready OS" is changing the way I use computers… Today I'm releasing the first alpha of the OmaPilot plugin for Omarchy Linux and looking for more testers to iron out the Omarchy harness.

omarchy plugin add omarchy plugin validate ~/.config/omarchy/plugins/io.github.spencerbull.omapilot omarchy plugin enable io.github.spencerbull.omapilot right

Demo Specs: - @Dell XPS 14 with @intel Ultra X7 358H and Arc B390 - @Dell Pro Max Tower T2 with @intel Ultra 9 285K & @nvidia RTX PRO 6000 Blackwell connected through @Tailscale running @MiaAI_lab blueprint for Qwen3.8 27B DFlash2+NVFP4 - TTS via voxtype by @peteonrails ships by default with Omarchy on iGPU; runs whisper large-v3-turbo locally on the XPS 14 - @ElevenLabs for TTS - @OmarchyLinux as the agent-ready OS

Harness Features - Email & Calendar integrations with @37signals hey-cli - Native Omarchy and @hyprwm skills - Use your @OpenAI or @xai subscriptions Recommended models - Self-hosted Qwen3.6 27B offloaded via Tailscale - GPT-5.3 Codex Spark (yes, we found a use for it) - grok 4.6

@OmarchyLinux did the same thing but with realtime

@OmarchyLinux smooth work sailor. psyching to give it a go and find where we might contribute

@OmarchyLinux

@OmarchyLinux It's finally all starting to click for me, the only thing I see as a major issues is the lag.

@OmarchyLinux 2 weeks ago we didn't have Opus 4.6 level intelligence running locally at 300+ tps....give it a few weeks....also a good use for these new 'fast' models being released by the providers.

@OmarchyLinux Looks like what Apple should have done with Siri on macOS

@OmarchyLinux DAMMMM, THISSS LOOKSSS CRAZZYYYY GOODDDDDD

@OmarchyLinux Can you get your Aussie agent to say noaaarrrr

@OmarchyLinux The Aussie voice is sending me. Why does it sound like my corporate HR training? 😭

Hi @SpencerGBull, I got it working on my Omarchy Quattro on my old MSI Prestige 15. I'm running is with the Build-in OmaPilot (PI) harness and GPT-5.6 Luna since GPT-5.3 spark is available on my ChatGPT Plus subscription. Do you plan to add a configuration for toggling agent responses visibility, glow animation visibility or hiding caption text? Happy to contribute if you'd like to.

@OmarchyLinux Happy to join the effort as a tester.

@OmarchyLinux Amazing, installed it, made some tweaks and its live! - this is next level.

@OmarchyLinux Remove talk back for action tasks

@OmarchyLinux you can! in settings 😁

@OmarchyLinux I wrote an omaharness a few days ago. Works great, like yours:

@OmarchyLinux this is super cool. i almost glanced right over it because i don't have a machine capable of self-hosting llms with respectable speeds...but now i'm seeing i can use a hosted api.

@OmarchyLinux Amazing stuff. This is the future.

This is really sweet Spencer. Absolutely killer. I installed and am using it -- was able to get it to work, but having some trouble getting it to reply consistently not sure why. Got it to reply a few times though and it was real impressive. GPT has a realtime model that can considerably reduce the delay by consolidating both the input & output steps clip from voxtype -> reasoning model -> elevenlabs -> me

@OmarchyLinux Wait till Microsoft sees this, lol! Great work!!

@OmarchyLinux @davidfano check this out!

@OmarchyLinux Working great with kokoro but I had to install a shim to get around the 3s load timeout on the probe. I keep a warm kokoro HTTP service running and the shim proxies to it from the command line. Voxtype 1.0.0 improvement is imminent and I'll send a couple PRs!

@OmarchyLinux This is pretty sweet

@OmarchyLinux 🤤🤤🤤

@OmarchyLinux Wow this is the future of computing or was meant to be ...

@OmarchyLinux Nice. Will try it out later today. I have so many things I want from this!

@OmarchyLinux Insane. Imagine this for someone who has accessibility needs (Parkinson’s, quadriplegics, and so much more) Great stuff man

@OmarchyLinux I will put this right in front of my parents.

@OmarchyLinux 🔥 amazing

@OmarchyLinux I’ll take two them. Prod and Sandbox. You’re the best.

@OmarchyLinux Ps nice theme

@OmarchyLinux Did something similar with codex voice through the codex plugin, mostly controling luvus session and everything inside it like tab, pane and agents. give it try

@OmarchyLinux I’m curious how the “focus” works while this is running. Can you interact with it and let it do things in the background while you’re working on other things? Suppose I’m coding and I want it to open up some docs on the side for example

@OmarchyLinux Make the voice a female Aussie.

@OmarchyLinux .@elonmusk Once the long replies and pauses are removed, it's basically Iron Man tech. At this point, it should be easier to convert developers on Mac/Windows to Linux (especially Omarchy), than to convert CC/Codex users to Grok Build.

@OmarchyLinux @SpencerGBull PErché per me non funziona così? io non ascolto le sue risposte.,... Ho configurato ElevenLabs ApiKey

@OmarchyLinux

This looks so good. I have been pushing on this for a while, but this looks so clean and looks like it works amazing, gonna give it a go. As im getting older my wrists are getting bad so i want to have a purely voice driven OS that maybe also uses head tracking so i can finally free my hands from typing and mousing. Thank you for your efforts, cant wait to try it!