Prompting for Your Project
From Practice to Real Work
For the last four lessons, you've been building prompting skills through exercises and examples. In this lesson, you apply all of it to your actual project. This isn't practice — this is real building.
By the end of this lesson, you'll have a detailed project plan created through a series of AI-assisted conversations. This plan becomes your roadmap for Modules 3–7. It's the document you'll reference every time you sit down to build.
What your project plan should include:
- A clear one-paragraph project description (you wrote this in Lesson 1.4)
- A feature list: what your project will do (prioritized from essential to nice-to-have)
- A technical approach: which tools you'll use and why
- A rough structure: what are the main components/pages/sections?
- A milestone timeline: what you'll build in each module of this course
- Identified risks: what could go wrong and how you'll handle it
You're going to create each section with AI assistance — using the prompting techniques you've learned.
Answer: Lessons 2.1–2.4 taught techniques through exercises. This lesson is a working session where you apply those techniques to produce a real deliverable — your project plan. From here on, prompting is a tool you use, not a topic you study.
Your Prompting Workflow for Planning
Here's the sequence of conversations you'll have with AI to build your project plan. Each step uses a different prompting technique from this module:
Conversation 1: Feature Brainstorm
Techniques: Socratic + Command
Start with your project description from Lesson 1.4. Use Socratic prompting to explore features:
Sample Prompt:
"I'm building [your project description]. What features would make this genuinely useful for [your audience]? Think step by step about what the user would want to do and what problems they'd encounter. Then organize the features into three tiers: must-have, should-have, and nice-to-have."
After getting the initial list, use follow-up commands to refine: "Remove anything that requires a database. Add a feature for [specific thing you want]."
Conversation 2: Technical Approach
Techniques: Devil's Advocate
Describe your feature list and ask AI to help you choose tools:
Sample Prompt:
"Here are the features I want to build: [paste list]. I'm a beginner with no coding experience on Track [1/2/3]. Walk me through the best tools and approach for building this. Then play devil's advocate: what are the biggest risks with this approach?"
Conversation 3: Structure and Layout
Techniques: Few-Shot + Constraints
Use few-shot examples to get AI to map your project's structure:
Sample Prompt:
"You are a UX designer creating a project structure for a teen's first app. Here's the format I want: Page: Home | Purpose: Dashboard showing today's key info | Sections: Header, daily summary, quick actions | Key feature: [main thing user does here] | Now create this structure for my project: [description]. Keep it to 3–5 pages/screens maximum."
Conversation 4: Timeline
Techniques: Command + Constraints
Map your project to the course modules:
Sample Prompt:
"Map my project build to this 7-module timeline. I can work about [X] hours per week. For each module, tell me the specific milestone I should hit: M1-2: Planning (done) | M3: Core building | M4: Data structure | M5: Build, test, deploy | M6: Ethics review | M7: Polish and ship | My project: [description]. My features: [paste list]."
Answer: Devil's Advocate. "What's unrealistic about this timeline? Where am I trying to do too much? What will take longer than I think?" This is the same pattern from Lesson 2.3, now applied to your real project.
The Iteration Process in Action
As you work through these conversations, you'll notice something: AI's first response is never the final version. You'll iterate. Here's what productive iteration looks like during a real planning session:
- AI suggests 12 features. You realize 4 are unrealistic → "Remove features 3, 7, 9, and 11. They're too complex for a first project."
- AI's structure has 7 pages. Too many → "Consolidate this to 4 pages. Combine the settings and profile pages."
- AI's timeline is too aggressive for Module 3 → "I can only work 3 hours that week. Scale down the Module 3 milestone."
- AI's tool recommendation doesn't feel right → "I'm not sure whether to build my project as a single-page app or a multi-page site. What are the trade-offs for a [project type]?" (Socratic)
Each iteration makes your plan more realistic and more yours. AI is helping you think, but you're making the decisions. That's the builder's mindset in action.
Making Your Plan Real
After your conversations, compile everything into a single document. A project plan template is available in the activity below. Fill in each section with the best output from your AI conversations, edited in your own words.
Important: Don't just copy-paste AI output into your plan. Read it, revise it, and make it yours. If a sentence doesn't sound like something you'd actually say, rewrite it. Your plan should reflect your vision and your voice — AI is the assistant, you're the author.
When you're done, you'll have a 1–2 page document that's your project's North Star for the rest of the course. Every module from here on connects back to this plan.
Key Concepts
- A project plan includes: description, features, technical approach, structure, timeline, and risks.
- Different prompting techniques suit different planning tasks: Socratic for exploring, commands for defining, few-shot for structuring, Devil's Advocate for risk assessment.
- Productive iteration means reviewing AI output critically and directing specific changes.
- Your plan should be in your own words. AI assists; you author.
- This plan becomes your roadmap for Modules 3–7.
Try It: Build Your Project Plan
This is your main working session for this lesson. You'll have a detailed project plan ready by the end.
1. Open the Project Plan Template
Scroll down to the interactive template below. It walks you through each section of your plan step by step.
2. Work through Conversations 1–4
Follow the workflow described in Content Block 2. Have each conversation with your AI tool and save the responses.
3. Extract and compile
After each conversation, extract the useful output and paste it into the relevant section of your template.
4. Edit in your own voice
Go through everything you've pasted. Rewrite sections that don't sound like you. Make it personal.
5. Run final Devil's Advocate pass
Share your complete plan with AI and ask: "What are the three biggest weaknesses in this plan?" Make final adjustments based on the response.
6. Save and celebrate
You now have a real project plan. This is what you'll build from for the rest of the course.
Estimated time: 20–30 minutes of active AI conversation plus 10–15 minutes of editing.
Pro tip: Log your best prompts in your Prompt Iteration Tracker (PDF). These become reference material for future projects.
Find this and all other downloadable resources on the Dashboard Resources page.
Check Your Understanding
1. What's the purpose of a project plan?
Explanation: Your project plan is a living roadmap. It guides your work, keeps you focused, and gives you something to reference when you're deciding what to build next. It will evolve — that's normal and expected.
2. Which prompting technique is best for the "feature brainstorm" phase of planning?
Explanation: Start with Socratic questions to explore what features would be genuinely useful, then use commands to organize and refine the list. This combination produces a thoughtful feature list rather than a generic one.
3. You've compiled your plan from AI conversations. What's the most important next step?
Explanation: AI output is a starting point. Your plan needs your voice, your judgment, and your verification. Rewriting in your own words also forces you to truly understand every part of the plan — if you can't explain it in your own words, you don't understand it well enough.
4. Your AI-generated plan has 15 features and a 3-week timeline. What should you do?
Explanation: 15 features in 3 weeks is almost certainly unrealistic for a first project. Devil's Advocate prompting will help you identify which features are essential and which can be cut. Simple and complete beats ambitious and unfinished.
Reflect & Write
Write 3–4 sentences: How did it feel to use prompting techniques for real project work instead of exercises? What was the most useful prompting technique for your planning process? What surprised you about the AI's suggestions for your project?
Project Checkpoint — Major Milestone!
Deliverable: Your completed project plan.
Your plan should include:
- Project description (1 paragraph)
- Feature list (prioritized: must-have, should-have, nice-to-have)
- Technical approach (tools and why)
- Project structure (pages/sections/components)
- Module-by-module timeline
- Top 3 risks and how you'll handle them
Verification check: Have you rewritten AI output in your own words? Have you stress-tested the plan with Devil's Advocate prompting? Is the scope realistic for your available time?
Next steps: Save this plan. Print it if you want. You'll reference it in every module from here on.
Helpful Resource
Prompting Cheat Sheet (PDF) — A printable quick-reference guide covering all the prompting techniques from Module 2. Keep it handy as you build your project.
Find this and all other resources on the Dashboard Resources page.
Level Up: Coming Next
Module 3 — Building the Core. Your project plan is complete. Now you start building. In Module 3, you'll choose your track, set up your environment, and start creating the first working version of your project. Everything you've learned about AI and prompting becomes your tool for building faster and smarter.
Continue to Module 3 →