HT Sherpa

An agent is an AI helper with a job and rules. These lessons help you write its plan. Choose a step in the lesson map. Writing a plan does not turn on tools or grant access.

New words, in plain language
Local / cloud
Local runs on your computer. Cloud runs on computers managed online.
Model / tokens
A model produces AI answers. Tokens are small pieces of text; some services charge for the tokens used.
Account / identity
An account is a sign-in. Your cloud identity is the account you use for online services.
Connector / permissions
A connector links an AI tool to another service. Permissions are the access you allow.

Hour 01 of 04 · About an hour

Understand it

What an agent is. What it is not. A page of principles in your words.

When you finish: A principles page you can hand to any later agent.

A mountain path in six steps, from a conversation to a model on your own computer
A workflow is a fixed path. An agent chooses the next step.

Start with the work

A language model answers a question. Give it tools, a memory, and a mandate, and it can take a step, look at what happened, and take another. If you can write the steps on a card, write a workflow. Save judgment for work that will not sit still.

Principles before machinery

Write a sentence: this agent exists so that a specific person is better served in a specific way. If the sentence cannot be written, do not build the agent. Trust is earned. Word and deed stay in view. Consult a person before anything hard to reverse.

This assists your judgment. It does not replace licensed professional judgment. It is architecture, not a certification.