Lesson Content

The goal is not to memorize every command. The goal is to know what kind of control you need at the moment.

Use this command decision map:

1. Need to change quality, speed, or depth? Look for model, reasoning, fast, or plan controls.
2. Need to confirm where work is happening? Look for project, local, cloud, worktree, or status controls.
3. Need to connect or inspect tools? Look for MCP, skills, plugins, or app-related controls.
4. Need to review code or changes? Look for review mode.
5. Need to organize the chat? Look for fork, side chat, new chat, pin, compact, or memories.
6. Need to report a product issue? Look for feedback, and remove sensitive details first.
7. Need to approve something? Stop and understand the action before approving.

Professional Use Case: A course creator asks ChatGPT to produce a lesson, then asks Codex to update a WordPress course bundle. The creator uses normal prompts for content drafting, session commands for model/reasoning choices, project commands for the correct repo, review mode for checking changes, and status checks before a long run.

Student task: Build a one-page command decision map for three real situations you expect to face: learning, writing, and technical/project work. For each situation, list the command category you would check, a normal prompt you would type, and the review step before using the output.

?A command decision map is a simple cheat sheet you create for yourself. Example: "Learning a hard topic | Plan first; ask for examples; save notes; quiz myself." You are practicing how to choose controls, not proving that your account has every command.

Starter scenario: "I need ChatGPT to teach me a new tool, help draft a course lesson, and then help check a WordPress plugin update."

Starter clarification prompt: "Turn my three goals into a command decision map. For each goal, tell me which commands or command categories to check, what normal prompt to use, and what review step protects me from bad output."

Troubleshooting: If a command is missing, write the equivalent plain-language prompt. If the AI suggests a command that your interface does not show, ask it for a fallback. If the task touches money, customer data, production files, or health/legal/safety topics, add human review before acting.

Quality Rubric:

• The map is practical enough to use during real work.
• Each situation includes a review step.
• The student can explain what they would do if a command is unavailable.