I was just wondering if anybody here uses opensource/free systems such as Hermes hosted locally for their agents. I was just trying to spin up an os using Hermes, wrapped in Ollama, hosted locally, with obsidian for memory, all hosted on a macbook pro.

🎉 Thanks to @Nate Herk 🎉 I have a wonderful new app that needs Beta testers!! What it does: I built the 28 Day Action Plan app as a 28 Day roadmap for people who want to accomplish a specific goal in 28 days, or, are working towards a much bigger goal that needs to be accomplished through focused monthly sprints. My proprietary framework has resulted in successful outcomes in: complex apps successfully built, whole house decluttering, fitness plans adhered to, financial stability achieved through daily focused efforts to track financial health, businesses started, and more. To sign up to be a beta tester please visit this thread. https://www.skool.com/28-day-action-plan-6088/updated-28-day-action-plan-app-needs-testers?p=5dbe69b2

Not trying to be negative, genuinely asking. cause I see a ton of people building them (myself included) but barely anyone talking about actually closing a client on one. single automations? easy sell. "this follows up with your leads automatically." done. business owner gets it in 10 seconds. but try explaining an AI OS to someone whos still running everything off spreadsheets and email lol. they dont want a "system" they just want to stop losing leads or stop drowning in admin work every morning. so whats actually working here? are yall selling the OS as a product? or just building it under the hood and selling the outcomes on top? would love to hear from anyone whos actually closed one of these 🙏 or anyone struggling w the same thing honestly. lets figure this out together

I'm running two businesses out of Chesapeake, VA. One is Platypus Automations, an AI automation agency serving local trade businesses in Hampton Roads. The other is a government contract brokering operation where I source federal work and sub it out. Right now my workflow is all over the place. Claude Code open, n8n open, Google Sheets open, session logs, prospect notes, pipeline data. I'm copying and pasting constantly just to keep things connected. What I want to build is one AI OS that ties all of it together. Morning briefing, pipeline status, what to work on, what's overdue, prospect research, outreach drafts, all from one place without touching five different tabs. Anyone here built something similar across multiple business lines? Curious what your setup looks like before I dive into the Level 3 content.

I'm a visual person. Always have been. A few months ago I set up Nate's Executive Assistant + layered Karpathy's wiki-style brain on top. Without really realizing it, that combo has been quietly evolving into my own AI OS — every skill, every workflow, every reference, all interconnected. Today I saw a post about HTML artifacts in agent workflows. Something clicked. Almost anything I run in Claude Code can be turned into a clickable, locally-hosted HTML page. A pitch deck. A project status overview. A deep research output with filters. An ingest flow with checkboxes per proposed wiki edit. A daily focus dashboard. Whatever the skill produces — text in chat OR a visual surface I can actually interact with. I started with my Daily Focus skill. It already: — scans projects, sessions, tasks, _hot.md — ranks by revenue proximity — reads my calendar for Maker Time blocks — outputs today's execution list Now it also writes a local HTML where I check off my floor + 3 build picks, add notes, see my streaks. I open it in VS Code's Live Preview — auto-refreshes whenever the file changes. The clever bit (or hack, depending on how you see it): no real two-way sync. A "Copy state" button at the bottom puts every change as JSON on the clipboard. I paste it in Claude, it updates all the relevant markdown files. Poor man's sync — nothing to maintain, no server, works offline. I built an ingest dashboard the same way. When Donna processes 25 raw files in 5 batches, each batch becomes a card with checkboxes per proposed wiki edit. Three states: approve, revise (with a required note), or skip. Click through, paste the state back, she applies only what I approved. Way more scannable than a wall of text in chat. Plus a Command Center index.html that lists every dashboard with status, last-updated time, and quick triggers. One landing for everything. I've been deep in this for hours and might keep going. Curious how others are using HTML in their workflows — anyone else stumble onto this?