AI Kick StartBook strategy call

Lesson 01 of 11 · Module 1

Foundation: AI tools landscape, accounts and MCP

Give learners the map: which products exist, what surfaces they run on, how accounts and plans affect capability, and why MCP is a permissions boundary.

Built from aikickstart_sec01.md, research/course_insight.md and screenshot catalogues

Next lesson

This module turns the source pack's broad research into the course operating map. The learner should leave able to explain the difference between chat assistants, coding agents, desktop agents, connectors, MCP servers, skills and plugins in plain English.

What to understand

  • Start with jobs to be done, not brand preference. ChatGPT, Codex, Claude, Claude Code and Cowork overlap, but they are not interchangeable.
  • MCP is an integration protocol and a trust boundary. A server can expose useful data and actions, so learners need approval, source and scope habits early.
  • Account and plan choices affect feature access. Treat specific prices and limits as volatile and re-check official docs before a paid workshop.
Visualisation

Tool surface decision map

Best first useRisk to manageLearner output
ChatGPTGeneral knowledge workMemory, uploads, connectorsPersonal workflow checklist
CodexProject and repo tasksSandbox, approvals, Git reviewSafe repo task brief
ClaudeLong-context reasoningSensitive uploads, connectorsProject setup rules
Claude CodeTerminal and IDE workPermission mode, shell accessCLI safety checklist
MCPTool integrationOAuth scopes, server trustLeast-privilege connector plan

Step by step

1

Build the course map

Build the course map - product screen reference

List every tool surface the learner will touch and classify it as chat, code, desktop automation, connector, MCP or extension.

HintUse plain roles: read, draft, edit, run, connect, schedule.

2

Mark permission boundaries

Mark permission boundaries - product screen reference

For each surface, mark what can read data, what can write data, and which action requires human approval.

HintApproval rules should be based on consequence, not confidence.

Reference screens

Course screenshots and visual references for the lesson flow. Re-check the live product before paid delivery or public launch.

Screen reference
Codex overview used to introduce agent surfaces.
Codex overview used to introduce agent surfaces.
Approval and safety reference for permission-first teaching.
Approval and safety reference for permission-first teaching.
Hands-on task

Create a one-page course map that shows tool, surface, best use, permission boundary and first learner output.

What you produce

A reusable AI platform decision map and account/setup checklist.

Production prompt examples

Tool surface classifier
Goal:
[What outcome should exist by the end of this lesson?]

Context:
[Audience, account tier, device, constraints, and current workflow.]

Inputs:
[Screens, docs, local files, or example data allowed for this exercise.]

Allowed actions:
[Read, draft, compare, summarise, or inspect.]

Ask before:
[Connecting apps, writing to files, sending externally, spending quota, changing settings.]

Output:
[The exact worksheet, plan, checklist, or capture pack to produce.]

Definition of done:
[How the learner or facilitator checks the result.]

Start by restating the plan in five bullets before executing.

Common mistakes to avoid

  • Teaching MCP as a technical side topic instead of a security and permissions topic.
  • Letting exact pricing or feature-limit claims ship without a dated source review.

Key terms

MCP
A standard for connecting AI tools to external data and actions.
Surface
The place a learner interacts with an AI tool: web, desktop, mobile, CLI, IDE or GitHub.

Resources

Checkpoint

Which tool surface should a non-technical learner use first, and what is the first approval boundary they must understand?