One Agent, Many Devices? 🧠Full Hermes Agentic AI + Tailscale Workflow
This tutorial demonstrates how to access a single Hermes AI agent from multiple devices using Tailscale VPN and a remote dashboard setup, enabling unified session history, centralized file access, and continuous memory across laptop and desktop without fragmentation.
Summary
Callum (Waterloots) provides a comprehensive guide for setting up a multi-device Hermes AI agent system. The core concept is maintaining one unified AI agent accessible from multiple devices—desktop and laptop—rather than running separate instances on each device. The key advantages presented include: consolidating session history across all devices into one source of truth, leveraging the more powerful device (desktop with 32GB RAM) to run larger local models while accessing them from a portable device (laptop with 24GB RAM), maintaining a centralized file system rather than fracturing files across devices, allowing the desktop to run 24/7 while the laptop can power down, and implementing end-to-end encryption through Tailscale so no data touches cloud servers. The setup process involves three main steps: installing Tailscale on both devices to create a secure private VPN tunnel between them, configuring dashboard credentials (username/password/secret) in the Hermes environment file for authentication security, and starting the remote dashboard with the command 'hermes dashboard-host 0.0.0.0' which makes it available on the local network at port 9119. For laptop access, users can either open the web dashboard in a browser using the desktop's Tailscale IP address, or configure the native Hermes desktop app to use a remote gateway connection. The web dashboard approach is simpler and supports non-default profiles, while the desktop app requires using the default profile due to a known limitation. For file sharing between devices, Callum demonstrates mounting the Hermes workspace folder across Tailscale using macOS file sharing settings (System Settings > Sharing > File Sharing) and connecting via SMB protocol, allowing the same workspace to appear on both devices. This is particularly important for Docker container setups where mounted volumes enable the agent to access files added from either device. The transcript emphasizes that all session data, chat history, and memories are stored on the desktop backend, so the laptop functions as a 'window' into the desktop's processing power and memory. Callum notes a potential limitation: if the desktop backend stops running, the laptop may revert to a local Hermes instance with separate session history and memories unless a separate memory sync strategy is implemented. He concludes by previewing future content on memory providers (Hindsight, Honcho, Nemesis) and building a tiered memory stack that compounds over time.
Key Insights
- Running a unified agent across devices prevents fragmented conversations in context—all messaging systems (Telegram, Discord, email) write to one centralized brain rather than creating separate session histories on different devices
- The desktop runs the actual Hermes backend while the laptop functions purely as a window or client interface; the chat history and session data remain stored only on the desktop, consolidating memory in one physical location
- Tailscale provides end-to-end encryption between devices without requiring cloud servers—it creates a private tunnel that works identically whether devices are on the same WiFi network or the user is traveling abroad
- The desktop app remote gateway setup currently has a limitation: it only works reliably with the default profile, and using non-default profiles causes a known issue where the app cannot reconnect to session history after restart
- File sharing across devices via SMB mounting is essential for Docker container setups because the agent cannot access files added from the laptop client unless they're mounted to the same workspace that the backend can see
Topics
Transcript
[0:00] I just asked a question and it ran on my desktop and then sent the answer back to my laptop. My laptop was just the window to my desktop which actually did the work. Whether it's across the room or across the world, I don't have two different session history and two different memory systems being built here. It's all being built on one device. Hi, my name is Callum, also known as Waterloots, and welcome to today's video on accessing your Hermes agent from multiple devices. This video assumes that you already have Hermes set up, but if you don't, I have a dedicated setup video and a full playlist that [0:31] helps you go deeper into the…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Wanderloots
Safer Agentic AI? 🧠Full Docker + Hermes Agent Tutorial (Desktop)
This tutorial demonstrates how to safely run AI agents like Hermes inside Docker containers to sandbox code execution and prevent unauthorized access to sensitive files on your computer. By configuring Docker volumes and mounting specific folders, users can give agents access to only the files they choose while keeping code execution isolated from the main system.
Full Hermes Agent Tutorial (Desktop) 🧠A Useful Agentic AI Workflow
This tutorial demonstrates how to set up and use Hermes, an open-source AI agent by Nous Research that learns and evolves over time through persistent memory and automated skill generation. The video covers local model setup with Ollama, cloud model integration with OpenAI, messaging gateway configuration via Telegram, and creating a self-improving daily AI briefing automation.
100% Free & Private AI 🦙 Build & Run Local AI Agents #ai #agenticai #ollama #localllm
The video introduces local AI models as a privacy-preserving alternative to cloud-based AI services. By running AI assistants locally on your computer, users can maintain complete control over their data while leveraging AI capabilities for personal tasks like analyzing notes and answering work-related questions.
Information ➡️ Knowledge: How To Build An LLM Wiki In Obsidian 🧠#obsidian #ai #agenticai
This video introduces the concept of an 'LLM Wiki' built in Obsidian — a structured knowledge system where AI agents automatically extract and organize concepts from raw information. The presenter explains that the goal is to transform information into persistent, accessible knowledge. This episode focuses on the practical setup, following a prior video that covered the rationale.
Obsidian Smart Plugin Workflow đź“ť New Smart Connections + Context AI
Callum (Waterloo Loot) demonstrates a three-plugin 'smart loop' workflow in Obsidian using Smart Connections, Smart Context, and Smart Chat to solve the problem of losing notes and context over time. The workflow follows a discovery-preparation-use cycle that helps users surface relevant notes, bundle them into reusable context, and optionally pass that context to AI while keeping chat threads linked to specific projects. The entire system is built on a shared 'smart environment' vector database layer that automatically updates context bundles as notes evolve.