Introduction to Prompt Modes
Most people interact with AI assistants in exactly one way: they type a question and hope for the best. Sometimes it works. Often it doesn’t. And they’re never quite sure why.
The problem isn’t the AI. The problem is using a hammer for every job—including the ones that need a screwdriver.
This course introduces you to six distinct prompt modes, each designed for specific types of tasks. Master these modes, and you’ll transform from someone who uses AI to someone who collaborates with it.
Why Modes Matter
Think about how you communicate with humans. You don’t talk to your doctor the same way you talk to your friend. You don’t write an email the same way you write a text message. Context shapes communication.
AI is no different. The way you frame your prompt determines the type of response you get. A question gets an answer. A command gets execution. A vision gets strategic thinking.
flowchart TD
A[Your Task] --> B{What do you need?}
B -->|Quick answer| C[Ask Mode]
B -->|Change existing content| D[Edit Mode]
B -->|Execute complex task| E[Agent Mode]
B -->|Step-by-step roadmap| F[Plan Mode]
B -->|Analyze & decide| G[Strategy Mode]
B -->|Long-term thinking| H[Vision Mode]The Six Prompt Modes
1. Ask Mode
Purpose: Get information, explanations, or answers When to use: You need to know something Example: “What’s the difference between REST and GraphQL?”
Ask Mode is conversational. You’re having a dialogue. The AI responds to questions, explains concepts, and provides information.
2. Edit Mode
Purpose: Modify, improve, or transform existing content When to use: You have something that needs changing Example: “Make this email more professional” or “Refactor this function to use async/await”
Edit Mode takes input and produces modified output. You provide the raw material; AI shapes it.
3. Agent Mode
Purpose: Execute multi-step tasks autonomously When to use: Complex work that requires multiple actions Example: “Create a new React component with tests and documentation”
Agent Mode treats AI as a capable worker. You describe the outcome; AI figures out the steps.
4. Plan Mode
Purpose: Create structured roadmaps and action plans When to use: You need to organize complex work Example: “Create a 30-day plan to launch my new product”
Plan Mode produces organized, sequential thinking. It breaks big goals into manageable steps.
5. Strategy Mode
Purpose: Analyze situations and recommend decisions When to use: You face choices and need clarity Example: “Should I use PostgreSQL or MongoDB for this project? Here’s my context…”
Strategy Mode is analytical. It weighs options, considers tradeoffs, and recommends paths forward.
6. Vision Mode
Purpose: Explore long-term possibilities and big-picture thinking When to use: You’re shaping direction, not executing tasks Example: “What could my business look like in 5 years if we focus on AI integration?”
Vision Mode is expansive. It explores possibilities, challenges assumptions, and thinks beyond immediate constraints.
Mode Selection Framework
Choosing the right mode is half the battle. Use this decision guide:
| If you need to… | Use this mode |
|---|---|
| Learn or understand something | Ask |
| Improve existing content or code | Edit |
| Complete a multi-step task | Agent |
| Organize work into steps | Plan |
| Make a decision between options | Strategy |
| Think about the future | Vision |
Common Mistakes
Using Ask Mode for Everything
Many people phrase everything as a question, even when they want action. “Can you write a blog post about X?” is less effective than “Write a blog post about X” (Edit/Agent mode).
Skipping Context
All modes work better with context. “Edit this” is vague. “Edit this email to be more concise and professional for a C-level audience” is specific.
Wrong Mode for the Task
Asking for a “plan” when you really need a “strategy” (or vice versa) produces misaligned output. Plans are about how. Strategies are about what and why.
Not Iterating
Prompts rarely work perfectly the first time. Each mode supports follow-up. Refine, redirect, dig deeper.
How This Course Works
Each subsequent lesson focuses on one mode in depth. You’ll learn:
- When to use the mode (ideal scenarios)
- How to structure prompts for that mode
- Examples across different use cases
- Common pitfalls and how to avoid them
- Practice exercises to build skill
By the end, you’ll instinctively know which mode to use and how to craft prompts that get results.
Your First Multi-Mode Exercise
Let’s see the modes in action with a single project: launching a newsletter.
Vision Mode: “What could a successful newsletter about productivity for developers look like in 2 years? Explore possibilities.”
Strategy Mode: “Given these options for newsletter platforms—Substack, ConvertKit, Beehiiv—which would you recommend for a technical audience? Here’s my situation…”
Plan Mode: “Create a 4-week launch plan for my developer productivity newsletter.”
Agent Mode: “Write the welcome email sequence for new subscribers (3 emails).”
Edit Mode: “Make this draft of my first newsletter more engaging and add a stronger call-to-action.”
Ask Mode: “What’s a good open rate benchmark for technical newsletters?”
Notice how each mode serves a different phase of the project. Vision explores. Strategy decides. Plan organizes. Agent executes. Edit refines. Ask fills gaps.
Key Takeaways
- Different tasks require different communication approaches
- Six modes cover the spectrum of AI interaction
- Mode selection is as important as prompt quality
- Each mode has specific structures and best practices
- Mastering modes transforms AI from tool to collaborator
Let’s dive into Ask Mode in the next lesson.