Complete AI Training
Sign inGet started

Video course · 13 chapters · 30 min · certificate

AI Agents for Beginners: From Chatbots to Autonomous Tools

AI Master teaches what separates agents from chatbots and how to connect them to tools, files, and apps. The course covers a hands-on demo, six agent tools, permission control, prompt contracts, and project memory.

Become a memberAlready a member? Sign in

What you'll learn

  • Understand the difference between a chatbot and an AI agent.
  • Identify the parts of an agent: model, tools, memory, and goal.
  • Connect agents to tools using MCP servers.
  • Compare agent tools like Claude Code, Codex, Antigravity, Grok Bot, Hermes Agent, and OpenClaw.
  • Set permission rules based on risk and reversibility.
  • Write a short prompt contract with boundaries and a done condition.

Chapters

13 chapters · 30:18
  1. 1:54 01Intro Members AI Agents Course Overview and Dinosaur Folder Demo Watch an agent turn a messy dinosaur folder into sorted, named, usable files.
  2. 3:36 02Foundations Members What AI Agents Are An agent is a model plus tools, memory, and a goal running in a loop.
  3. 2:56 03Tool protocol Members MCP: Giving Agents Tools MCP is a standard connector that lets one agent use many tools through small servers.
  4. 1:34 04Coding agent Members Claude Code Claude Code is a local agent that writes code, edits files, and tracks long work on your machine.
  5. 1:24 05Sandbox agent Members Codex Codex works in an isolated sandbox copy, so a bad run does not touch your original folder.
  6. 2:09 06Visual agent Members Antigravity Antigravity is a Gemini-built agent with visible task lists, browser control, and editor extensions.
  7. 1:57 07Bot fleet Members Grok Bot Grok Bot gives always-on bots on a dedicated machine that can share work and approvals.
  8. 1:51 08Async agent Members Hermes Agent Hermes Agent runs in the background, checks triggers, and reports to chat before making changes.
  9. 1:22 09Self-hosted Members OpenClaw OpenClaw is a self-hosted agent you control, from model choice to tools and behavior.
  10. 1:26 10Decision guide Members Choosing the Right Agent Start where the work lives, then match the agent to the job type.
  11. 1:41 11Permissions Members How Much Control to Give Match agent permissions to how easy an action is to reverse.
  12. 5:18 12Prompting Members The Prompt Contract Give agents a clear contract: goal, boundaries, input, output, and a done condition.
  13. 3:10 13Memory Members Agent Memory and Final Thoughts Use project memory to make agents better over time, and keep the file short.