Combining Modes: Real-World Workflows

Now that you understand each mode individually, let’s see how they work together. Real projects rarely fit into a single mode. The skill is knowing when to switch and how to chain them effectively.

The Mode Spectrum

flowchart LR
    A[Vision] --> B[Strategy]
    B --> C[Plan]
    C --> D[Agent]
    D --> E[Edit]
    E --> F[Ask]
    
    A --> |"What's possible?"| A
    B --> |"What's best?"| B
    C --> |"How to execute?"| C
    D --> |"Build it"| D
    E --> |"Improve it"| E
    F --> |"Clarify it"| F

The spectrum flows from expansive thinking (Vision) to focused refinement (Ask/Edit). Different phases of work call for different positions on this spectrum.

Mode Progression Patterns

Pattern 1: Full Project Lifecycle

Vision → Strategy → Plan → Agent → Edit → Ask

This is the complete flow for a major project:

1. Vision: “What could our new product look like?”
2. Strategy: “Of these possibilities, which should we pursue? How does option A compare to B?”
3. Plan: “Create a 12-week roadmap to build this”
4. Agent: “Create the authentication system with tests and documentation”
5. Edit: “Improve this onboarding flow copy to be more engaging”
6. Ask: “What’s the best practice for handling refresh tokens?”

Pattern 2: Decision Flow

Ask → Strategy → Plan

When you need to make a decision and then act:

1. Ask: “What are the differences between PostgreSQL and MongoDB?”
2. Strategy: “Given my specific context [details], which should I choose?”
3. Plan: “Create a migration plan to move our data to PostgreSQL”

Pattern 3: Iteration Flow

Agent → Ask → Edit → Agent

When building and refining:

1. Agent: “Create a landing page for our new feature”
2. Ask: “What are conversion rate best practices for SaaS landing pages?”
3. Edit: “Apply these principles to improve the landing page”
4. Agent: “Now add an A/B testing setup for the two variations”

Pattern 4: Problem Solving Flow

Ask → Vision → Strategy → Plan

When facing a challenge:

1. Ask: “Why might our user retention be dropping?”
2. Vision: “What are all the possible approaches to improving retention?”
3. Strategy: “Analyze which of these approaches would have the highest impact for our situation”
4. Plan: “Create an implementation plan for the top two strategies”

Real-World Workflow Examples

Example 1: Starting a New Product

Phase 1: Exploration (Vision)

We're a B2B SaaS company with expertise in workflow automation.
Our customers keep asking about AI integration.

What could an AI-powered workflow product look like in 2 years?
Explore different directions we could take.

Phase 2: Decision (Strategy)

Based on the possibilities we explored, I'm considering three directions:
A) AI assistant for workflow design
B) AI-powered workflow optimization
C) Natural language workflow creation

Given our team's capabilities [details], market position [details], 
and resources [details], analyze which direction is best.

Phase 3: Roadmap (Plan)

We've decided to pursue natural language workflow creation.
Create a 6-month product development plan including:
- MVP scope
- Technical milestones  
- Team allocation
- Go-to-market timeline

Phase 4: Execution (Agent)

Create the initial architecture document for the NL workflow engine.
Include:
- System components
- Data flow diagrams
- API design
- Integration points with existing product

Phase 5: Refinement (Edit)

Improve this architecture document to:
- Add more detail on the parsing engine
- Include error handling strategies
- Make it clearer for junior developers

Phase 6: Clarification (Ask)

The team has questions about the event-driven architecture.
Explain the benefits and tradeoffs of event sourcing 
for our workflow state management.

Example 2: Solving a Business Problem

Problem Identification (Ask)

Our free trial conversion rate dropped from 15% to 8% over 3 months.
What are the most common causes of conversion rate drops in B2B SaaS?

Root Cause Analysis (Strategy)

Here's our data:
- Trial starts unchanged
- Time to first value: increased from 2 days to 5 days  
- Feature adoption in trial: decreased
- No product changes in period

Help me diagnose the most likely root causes.
What additional data should I look at?

Solution Exploration (Vision)

Given that the root cause appears to be slower time-to-value,
what are all the possible ways we could accelerate the trial experience?
Think broadly—what do the best companies do?

Solution Selection (Strategy)

From the possibilities explored, evaluate these three approaches:
A) Guided onboarding wizard
B) Pre-populated demo data
C) Live onboarding calls for all trials

Our constraints: 2-week implementation window, 1 developer, no sales capacity.

Implementation Planning (Plan)

Create a 2-week implementation plan for the guided onboarding wizard.
Include daily tasks, testing approach, and rollout strategy.

Execution (Agent)

Create the onboarding wizard component including:
- Step-by-step flow (5 steps)
- Progress indicator
- Skip functionality  
- Completion tracking for analytics

Example 3: Technical Decision Making

Information Gathering (Ask)

Explain the tradeoffs between REST and GraphQL APIs.
Focus on:
- Development complexity
- Client flexibility
- Caching
- Performance at scale

Context-Specific Analysis (Strategy)

We're building a dashboard application with:
- Multiple widgets showing different data
- Real-time updates needed
- Mobile and web clients
- Team is experienced with REST, new to GraphQL

Given this context, should we use REST, GraphQL, or a hybrid?

Decision Made—Now Plan (Plan)

We've decided on GraphQL with REST fallback for simple endpoints.
Create an API design plan including:
- Schema structure
- Query organization
- Authentication integration
- Migration approach for existing REST endpoints

Mode Selection Quick Reference

If you’re…	Start with	Then move to
Exploring new territory	Vision	Strategy
Making a choice	Ask (for info) → Strategy	Plan
Executing a defined task	Agent	Edit (for refinement)
Improving existing work	Edit	Ask (for guidance)
Debugging/understanding	Ask	Strategy (if options)
Planning execution	Plan	Agent (to execute)
Stuck or overwhelmed	Vision (step back)	Strategy (prioritize)

Mode Switching Signals

Switch FROM Ask TO Strategy When:

• You have enough information
• You face a choice between options
• You need a recommendation, not more facts

Switch FROM Strategy TO Plan When:

• Decision is made
• You need to organize execution
• You want a roadmap

Switch FROM Plan TO Agent When:

• Plan is clear
• You’re ready to build
• Individual components need creating

Switch FROM Agent TO Edit When:

• Initial version exists
• Quality needs improvement
• Specific changes needed

Switch TO Vision When:

• You feel stuck
• Options seem limited
• You’re questioning the goal itself

Common Multi-Mode Mistakes

1. Skipping Modes

Jumping from Vision straight to Agent skips critical thinking. The output suffers.

2. Staying in One Mode

Using only Ask Mode when you need Agent Mode wastes effort. Using only Agent Mode when you need Strategy produces unfocused work.

3. Wrong Order

Planning before strategic decisions are made produces wasted plans. Executing before planning produces unfocused execution.

4. Not Returning to Earlier Modes

New information might require revisiting Vision or Strategy. Don’t treat the sequence as one-way.

Practice Exercise: Multi-Mode Project

Choose a real project you’re working on. Map out which modes you’d use at each phase:

1. Project Definition Write down your project goal.

2. Mode Mapping For each phase of your project, identify:

• Which mode fits best
• What prompt you’d use
• What output you expect
• What triggers the next mode

3. Create the First 3 Prompts Write actual prompts for the first three modes you’d use.

4. Identify Iteration Points Where might you need to loop back? What would trigger that?

Key Takeaways

• Real projects flow through multiple modes
• Vision → Strategy → Plan → Agent → Edit → Ask is the full spectrum
• Different project phases need different modes
• Watch for signals that it’s time to switch
• Don’t skip modes—each serves a purpose
• Be willing to loop back when new information emerges
• Mode mastery is about fluid transitions, not rigid rules

Course Summary

You’ve learned six powerful modes for AI communication:

Mode	Purpose	Key Question
Ask	Get information	”What do I need to know?”
Edit	Transform content	”How should this change?”
Agent	Delegate execution	”What needs to be built?”
Plan	Organize work	”How do I execute this?”
Strategy	Make decisions	”What should I do?”
Vision	Expand thinking	”What’s possible?”

The skill isn’t just knowing the modes—it’s knowing when to use each and how to flow between them.

Start practicing today. Pick a real project. Identify which mode you need. Write a prompt. See what happens. Iterate.

The more you practice, the more natural mode selection becomes. Eventually, you won’t think about modes consciously—you’ll simply communicate with AI effectively.

That’s the goal. Now get to work.

What's the typical mode progression for a major project?

Vision → Strategy → Plan → Agent → Edit → Ask

Ask → Edit → Agent → Plan → Strategy → Vision

Plan → Agent → Edit → Ask → Vision → Strategy

Agent → Edit → Ask only

When should you switch from Strategy Mode to Plan Mode?

When you need more information

When a decision is made and you need to organize execution

When the first version is complete

When you feel stuck

What's a signal that you should switch TO Vision Mode?

You have enough information to decide

You're ready to start building

You feel stuck or options seem limited

Quality needs improvement

What's a common multi-mode mistake?

Using multiple modes in one project

Starting with Vision Mode

Ending with Edit Mode

Skipping Strategy and jumping from Vision to Agent

Which statement best describes mode mastery?

Always using the same mode for consistency

Fluid transitions between modes based on project needs

Using as many modes as possible in every project

Only using Agent Mode for speed

Submit