← ClaudeAtlas

building-autonomous-agentslisted

The mental model behind an always-on system — a tool is something you invoke, an agent is something that runs, and the gap between them is a trigger. Teaches the Sense → Correlate → Judge → Act → Report loop every agent shares, the two ways to deploy it (a job scheduled on your machine vs. a hosted routine that runs 24/7 even with the laptop closed), the real gotchas that quietly break both, and the Observe → Propose → Act trust ladder that keeps it from doing damage while you sleep. Use when you keep hand-running the same job, when you want to understand what "agentic" actually means before you build, when a "set it and forget it" agent isn't firing on its own, or when you're deciding whether a job should run on your machine or in the cloud.
search-atlas-group/amm-founding-circle · ★ 0 · AI & Automation · score 72
Install: claude install-skill search-atlas-group/amm-founding-circle
# building-autonomous-agents **The one idea this skill exists to teach:** *a tool is something you invoke, an agent is something that runs.* The gap between them is a **trigger** — plus the discipline to trust it. Most of what people proudly call "agents" are tools with extra steps. You still open a terminal, run the thing, watch it, and remember to do it again tomorrow. That's a very good tool. It is not an agent. An agent has a trigger that fires *without you* — a schedule, an inbox that filled up, a form that got submitted — and it does its job whether or not you're looking. If nothing fires without you, it's not an agent yet. That's the whole test. This is the mental model, not a piece of software. Read it once and the rest of the always-on skills (`host-your-agent`, `night-shift`, and the runner behind them) stop feeling like magic and start feeling like plumbing you understand. --- ## Why this matters for an agency owner You are the trigger for almost everything in your business right now. The report gets pulled because *you* pull it. The client gets the update because *you* remember to send it. The inbox gets swept because *you* sit down and sweep it. Your capacity is capped at "things you personally got to today." An autonomous agent moves work off that list. Not by being smarter than you — by *running* when you're not there. The mindset shift is the one Boris Cherny names: **agents are capacity, not tools.** You stop thinking "what script do I run" and start