Hermes Agent has too many plugins. No idea which ones your agent needs?🥸 👉Hermes Agent makes choosing easier: explore plugins by category, understand what they do, and find the right fit for your workflow. Less guesswork. More building.🚀

Every time I think “I wish my computer could do this,” it turns into another little project 😂 This is why I love Omarchy.@dhh I can build my own widgets, bring apps into them, add tabs, and shape the whole desktop around how I work. Here’s what I’ve been building: my own agent workspace, with Claude, Codex, and Hermes in one place. My computer is starting to feel like something I made.

Stop putting everything into Hermes memory. Some information only needs to survive this session.🚀 So I gave every Hermes session another side: CHAT ↔ SCRATCHPAD 🔄 Flip. Note it. Flip back. Keep working. Not everything deserves memory.✋

You forgot your Omarchy login password ⁉️Now what ? No worries - Hermes Can fix it👇 -Press Ctrl + Alt + F2 -enter your user then enter again -Type Hermes if you had installed - talk to hermes agent and get it fixed. An OS had never been so easy to fix. Hats Off @dhh

Is Hermes an Agent OS -This is not the correct question ? 🚀🪽 It’s: “What is still missing before we can trust it like one?” Two months ago, I called Hermes an Agent Operating System. Today, I’d put five expectations behind that phrase: ✋Isolation: Give each agent only the files, credentials, and access its job needs. A job title is not a security boundary. 👀Visibility: Show me which agent is working, what it changed, and where it’s stuck—not just “task completed.” 👷‍♀️Control: Let me steer or stop the worker doing the job without restarting the whole team. ⚡️Recovery: Contain failures, preserve good work, and clearly distinguish what can be restored from what cannot be undone. 😎Ownership: Let me change models without losing my workflows, history, and settings. My operation shouldn’t depend on one model staying the best. These are standards to test, not guarantees to assume.

Installed Hermes Agent. Now what?🚀 Not sure where to start—or what to build with it? There’s a page showcasing what people are building with Hermes Agent. Explore their projects, see what sparks an idea, and find a starting point for your own. ✋You don’t need a groundbreaking idea. Start with one repetitive task you wish someone else would handle. Explore → pick a problem → build your version. What would you want your agent to take off your plate?⁉️

🪽 OpenAI just overtook Anthropic in spending on OpenRouter. Does your Hermes setup reflect that shift? I’m curious what Hermes users are actually running: • OpenAI • Claude • Other hosted models • Local models Which is your daily driver—and what made you choose it?

🧬 5. SKILL ROUTER Give Hermes Agent specialization. Don’t hand every task every capability. Give Hermes the goal. Then: GOAL → DISCOVER → SELECT → LOAD → EXECUTE → VERIFY Research job? Load research capabilities. Coding job? -Load coding capabilities. Verification job? -Load only what the verifier needs. Why: A 3-tool problem shouldn’t automatically become a 40-tool context problem.

🧠 4. AGENT FLIGHT RECORDER Give Hermes accountability. Record the important trail: • decisions • tool calls • file changes • agent handoffs • failures • recovery attempts Then you should be able to ask: -What changed?” -Which agent changed it?” -Why did you do that?” -Can I roll it back?” Why: Agent observability becomes more important as autonomy increases.

📂 3. WORKSPACE GOVERNOR Give Hermes organization. Before Hermes creates anything: PROJECT → PATH → CONFIRM → WRITE → REPORT Make it: • print the current directory • identify the correct project • confirm the write target • restrict allowed files • return every artifact path Why: No more final-v2-new-really-final.md hiding somewhere on your machine. 😅

⏱️ 2. CRON GUARDIAN Give Hermes a heartbeat. Don’t only monitor whether a scheduled job failed. Detect when it never started at all. EXPECTED → RAN? → VERIFY → RETRY/ALERT Track: • missed runs • failed runs • stale jobs • retry attempts • last successful execution Why: Error handling can’t catch a job that never executed.

🔐 1. PERMISSION BROKER Give Hermes boundaries. Hermes can prepare an action, but sensitive actions hit an approval gate. SEND → ASK DELETE → ASK DEPLOY → ASK CALL → ASK Why: Autonomy without permission boundaries becomes expensive to trust.

Stop Giving Hermes More Superpowers. Fix These 5 Problems First. 🪽 Everyone wants to make their agent more autonomous. 🚀 More tools. More skills. More permissions. More agents. But after looking at what agent builders actually struggle with, the biggest problems are much more boring: “Why did it do that?” “Why didn’t the cron run?” “Where did it put my files?” “What exactly changed?” “Why did it load all those tools?” So these are 5 Hermes skills/workflows I’d build first 👇

What if your Hermes Agent has a complete Library of all the things related to Hermes Agent ? 🚀🪽 You got stuck need help. Pick a book. 📕. It has : 1. Complete Hermes Agent Guide 2. Skills Books which can be directly installed. 3. List of plugins and click to activate. 🚀 And many more to be added. 🔥🌟

✋Had: wait for a charging dock that turns the phone into Hermes. 🫵Now: Hermes already prints the morning paper — calendar, mail, GitHub fails, weather — while I sleep. Hardware can wait. Ops can’t.🚀 1. Same dream, cheaper form factor: overnight bulletin > plastic dock. Hermes Dispatch, not brunch news. 2. That dock was “agent in my hand.” The paper is “agent already did the handoff.” I’ll take receipts over screen.

My Hermes Agent has a newspaper for me now. 🗞️🚀 The Hermes Agent Times: four pages covering my day, my agents’ work, ideas worth exploring, and life beyond the screen. 🚀Next: a morning printout and an Obsidian archive connecting each day’s ideas. Would you read a newspaper made by your own agent? # Give below prompt to your Hermes agent👇 @Teknium @NousResearch

This Hermes update looks small.🚀😁 but this is the Power of Open Source. For teams, it isn't.🪽 You can now keep a Hermes plugin inside a private GitHub repo and install it using GitHub credentials already configured on the machine. Think: 🔒 company/deployment-plugin 🔒 company/support-plugin 🔒 company/internal-tools Your company doesn't have to publish its agent capabilities publicly just to distribute them to its own Hermes agents. Private repo → authorized machine → Hermes plugin This is how internal agent tooling should work.