Summary
This tutorial by Sharbel A. covers a structured methodology for setting up the Hermes AI agent as a genuine daily operating system — not a chatbot for one-off tasks, but a persistent assistant with memory, scheduled jobs, browser access, sub-agents, and tool integrations. The core argument is that most users approach Hermes backwards, coming in with tasks rather than letting the agent first understand their actual working life.
The recommended first step is a “Hermes life audit” — a structured interview sequence where Hermes asks the user about recurring responsibilities, recurring decisions, bottlenecks, and content workflows before any automation is configured. The specific prompt used produces five outputs: a ranked list of the top 10 automatable tasks, the three easiest workflows to deploy immediately, the three highest-leverage workflows to build over the week, the tools and permissions each workflow requires, and a daily morning check-in prompt. Practical examples from a real session include client experience monitoring tied to a live dashboard, post-call follow-up automation, and a YouTube content pipeline that pulls competitor transcripts and surfaces filming recommendations autonomously.
From the audit, the video walks through selecting exactly three initial workflows — one easy, one recurring, one high-leverage — to avoid the common mistake of trying to automate everything at once. The creator also covers how to build a delegation boundary map that distinguishes tasks Hermes can run unsupervised, tasks requiring human approval before execution, and tasks that should stay manual entirely. Full prompts used throughout are shared for direct copy-paste use.
📺 Source: Sharbel A. · Published June 28, 2026
🏷️ Format: Tutorial Demo
