Learn to work as a team
in the age of AI.

1. 01

   AI Fluency: the foundations

   done ✓
   Anthropic Academy · AI Fluency: Framework & Foundations ↗

   The shared mental model your whole team runs on — what AI is, what it isn’t, and the 4E framework for working with it well: Effective, Efficient, Ethical, Safe.

   AI Fluency’s “4E” framework names four ways to work well with AI. Which set is it?

   Effective, Efficient, Ethical, SafeExplore, Explain, Execute, EvaluateEasy, Exact, Eager, EqualEngage, Expand, Extend, Exit

   What is AI Fluency really trying to give you?

   Every prompt trick memorizedA durable mental model for thinking with and collaborating with AIA way to replace your team with agentsOne vendor’s API, by heart

   Check answers Ask SE-1 to go deeper →
2. 02

   Capabilities & limits

   done ✓
   Anthropic Academy · AI Capabilities and Limitations ↗

   A team delegates well only when it knows what Claude can and can’t do. Learn the real edges, so you hand off the right work — and verify the rest.

   Why does knowing Claude’s limitations matter for a team?

   So you never use itSo you delegate the right tasks and verify the restSo you can blame the AIIt doesn’t

   Claude gives a confident answer to a high-stakes question. The reliable habit?

   Ship it immediatelyVerify it against a source or a testAssume it’s wrongAsk it to sound more confident

   Check answers Ask SE-1 to go deeper →
3. 03

   Prompting as a shared language

   done ✓
   Anthropic Academy · Working with the Claude API & Prompt Engineering ↗

   When everyone prompts consistently, the team moves as one. The fundamentals — clear instructions, context, structure — that make Claude reliable for all of you.

   A weak Claude response. The most reliable way to improve it?

   Give clearer instructions, context, and examplesRepeat the same prompt, louderAlways use fewer wordsNever give examples

   Why structure a prompt with headings or tags?

   It just looks tidyIt helps Claude parse intent and produce consistent outputThe API requires itIt slows the model down

   Check answers Ask SE-1 to go deeper →
4. 04

   Cowork with Claude

   done ✓
   Anthropic Academy · Introduction to Claude Cowork ↗

   The heart of it: Claude as a teammate inside the shared workflow — drafting, researching, and reviewing alongside you, not just answering one-off questions.

   To “cowork” with Claude means…

   One person uses Claude in secretClaude works alongside the team inside the shared workflowReplacing the team with ClaudeOnly ever using Claude for chat

   A sign a team is coworking with AI well?

   Nobody checks the outputsClaude is woven into real tasks with clear handoffs and reviewOnly the manager is allowed to touch itIt gets used about once a month

   Check answers Ask SE-1 to go deeper →
5. 05

   Build together with Claude Code

   done ✓
   Anthropic Academy · Claude Code 101 + Claude Code in Action ↗

   Ship as a team — context management, hooks, custom commands, subagents, and MCP to connect Claude to your tools. Where coordination meets the keyboard.

   Claude Code is best described as…

   A chatbot windowAn agentic coding tool that works in your codebase and terminalA plain text editorA search engine

   What is MCP (Model Context Protocol)?

   A pricing planAn open standard for connecting AI to your tools and dataA specific prompt formatThe name of a model

   Check answers Ask SE-1 to go deeper →
6. 06

   Coordinate like masters

   done ✓
   Super Engineer · SE-1 — the craft no course teaches

   The human layer the tools can’t hand you: making and keeping promises, naming breakdowns without blame, repairing trust, and knowing when work is truly done. This is what makes an AI-native team unstoppable.

   A breakdown hits mid-project. The AI-native move?

   Find who to blameName it cleanly, renegotiate the promise, repair the trustPretend it didn’t happenEscalate straight to the CEO

   When is a piece of work actually “done”?

   When you’re tired of itWhen the agreed conditions of satisfaction are met and declaredThe moment the code compilesNever, really

   Check answers Ask SE-1 to go deeper →

Lessons live at Anthropic's Claude Academy · the quizzes, the trajectory, and SE-1 are Super Engineer · a working prototype