From Prompt to Pages
AI Prompt Guide for Building Your Canvas Course
This guide contains the prompt sequence used in the From Prompt to Pages workshop. The prompts are designed to work together. Each stage builds on decisions made and approved during the previous stage.
The Process
Use the Course Design Decision Guide to determine what you want.
Give AI your decisions, course materials, and instructions.
Evaluate the response rather than automatically accepting it.
Revise until the output reflects your pedagogy and intentions.
Carry the approved decisions forward into the next prompt.
Do not start a new AI conversation for every prompt. Keeping the prompts in the same conversation allows your approved course context, architecture, visual language, and design rules to accumulate.
Think. Approve. Build. Test.
Prompt 1: Create the Course Design Brief
Purpose
Before asking AI to design modules or Canvas pages, teach it about the course.
This prompt establishes the instructional context that should guide everything that follows.
What You Need
- Course title and discipline
- Course level
- Course modality
- Course length
- Official course description
- Student learning outcomes
- Major assignments
- Student experience or prior knowledge
- Your instructional priorities
- Areas where students typically need support
I am designing a course and want you to serve as an instructional design partner. I will remain responsible for the pedagogical decisions. Your role is to help me analyze, organize, design, and eventually build the course based on the decisions I approve. Here is the context for my course. Course title and discipline: Course level: Course modality: Course length: Official course description: Student learning outcomes: Major assignments: What students generally know or can do when they enter: Areas where students are likely to need additional guidance or scaffolding: My instructional priorities: The experience and tone I want the course to have: Do not design the course yet. First, create a concise Course Design Brief that identifies: 1. The primary instructional needs of these students. 2. The appropriate level of scaffolding. 3. Likely navigation and organizational needs. 4. An appropriate instructional tone. 5. Potential areas of cognitive overload or confusion. 6. Design priorities for the Canvas environment. 7. Any important tensions or challenges I should consider when designing the course. Separate conclusions based directly on the information I provided from assumptions you are making. Do not make consequential instructional decisions based on assumptions. Identify those assumptions and ask me to resolve them.
Review Before Continuing
- Does the brief accurately describe your students?
- Does it reflect the level of the course?
- Does it sound like your teaching?
- Did AI make assumptions you need to correct?
- Is anything important missing?
Do not continue until the brief reflects the course you intend to teach.
Prompt 2: Create the Course Architecture
Purpose
Now determine how learning should unfold across the course.
AI is helping organize the intellectual journey, not merely dividing content into weeks.
What You Need
- Your approved Course Design Brief
- Course learning outcomes
- Major assignments
- Required topics, readings, or content
- Any required institutional or departmental structure
- Your preferred organizing logic
Using the Course Design Brief we established and the decisions I have approved, help me develop the learning architecture for this course. Do not change the approved Course Design Brief unless I explicitly ask you to revise it. The course should be organized primarily around: Here are required topics, content, readings, activities, or other elements that must be incorporated: Here are any additional sequencing requirements or constraints: Develop a course architecture that shows how learning progresses from the beginning of the course through students' final demonstration of learning. For each module or major learning unit, identify: 1. Module title. 2. Central question, problem, or purpose. 3. Relevant student learning outcomes. 4. Essential content. 5. Learning activities or opportunities for practice. 6. Major assignment or deliverable, when applicable. 7. How this module builds on previous learning. 8. How it prepares students for what follows. Prioritize intellectual progression and learning relationships rather than simply distributing content evenly across the semester. Identify any places where the proposed sequence creates unnecessary cognitive load, assumes knowledge students may not yet have, or fails to adequately prepare students for a major assignment. Do not create Canvas pages or HTML yet.
Review Before Continuing
- Does the sequence make intellectual sense?
- Does each module have a purpose?
- Are assignments appropriately scaffolded?
- Can you see how students move toward greater independence or mastery?
- Did AI add content that does not belong in your course?
Prompt 3: Create the Visual Design System
Purpose
Now translate the learning architecture into a consistent visual language.
The goal is not decoration. The goal is recognition, hierarchy, navigation, and reduced cognitive load.
What You Need
- Your approved Course Design Brief
- Your approved Course Architecture
- The recurring types of information students encounter
- The terminology you want to use consistently
Using the approved Course Design Brief and Course Architecture, help me develop a consistent visual design system for the Canvas course. Do not select colors yet. The purpose of the visual system is to help students recognize the function and importance of information, navigate efficiently, and encounter a predictable course structure. The recurring types of information in my course include: The terminology I want students to encounter consistently includes: Develop visual treatments for the recurring course components. Consider components such as: Learning objectives or guiding questions. Required readings, videos, and other course materials. Learning activities and practice. Assignments and submissions. Important information and deadlines. Examples. Optional or supplemental materials. Help and support. For each component, recommend: 1. A consistent label or heading convention. 2. Its relative position in the information hierarchy. 3. An appropriate visual treatment. 4. How it should be distinguished from other types of information. 5. What should remain consistent every time the component appears. Prioritize clarity, accessibility, hierarchy, predictable navigation, and reduced cognitive load. Avoid decorative elements that do not serve an instructional or navigational purpose. Do not generate HTML yet.
Review Before Continuing
- Can students recognize recurring functions?
- Is the system simple enough to repeat?
- Does the terminology sound appropriate for your students?
- Is anything visually emphasized that does not deserve emphasis?
- Has AI added unnecessary decoration?
Prompt 4: Create the Color Logic
Purpose
Give color an instructional function.
Instead of selecting colors because they look attractive together, establish what each color means and when it should appear.
What You Need
- Your approved Visual Design System
- Any existing university, department, program, or course colors you want to use
- Your accessibility requirements
Using the visual design system we established, develop an accessible and consistent color logic for this Canvas course. Color must communicate function rather than simply decorate the page. My existing course, university, department, or program colors that I may want incorporated are: Create a limited palette for these functions: Course Identity Action Information Attention Support Neutral For each function, provide: 1. The recommended color. 2. The HEX code. 3. The instructional meaning of the color. 4. When the color should be used. 5. When the color should not be used. 6. Recommended text color when used as a background. 7. Any accessibility or contrast concerns. The palette should work as a coherent visual system. Do not rely on color alone to communicate meaning. Labels, headings, text, borders, or other visual cues must reinforce the function. Do not introduce additional colors without explaining why they are necessary. Do not generate HTML yet. Conclude with a concise Course Color Style Guide that can be reused in future prompts.
Review Before Continuing
- Does every color have a job?
- Are there too many colors?
- Can you explain what each color means?
- Will the same color always communicate the same function?
- Is the system accessible without color?
Prompt 5: Create the Page Architecture
Purpose
Now decide how recurring Canvas pages should function.
The goal is to build reusable page types rather than designing every page individually.
What You Need
- Your approved Course Design Brief
- Course Architecture
- Visual Design System
- Color Logic
- Your decisions about what students need from each page type
Using all of the course design decisions we have established and I have approved, create a reusable page architecture for this Canvas course. Do not change the established visual language or color logic. Create structures for these three page types: Start Here or Home Page Module Overview Page Assignment Page For each page type, identify: 1. Primary instructional purpose. 2. Information students need immediately. 3. Recommended content sections. 4. Information hierarchy. 5. Reusable visual components. 6. Primary student call to action. 7. Navigation to or from the page. 8. Elements that should remain consistent every time this page type appears. 9. Elements that can change according to the content. Design these pages as parts of one coherent course system rather than three unrelated pages. Do not generate HTML yet.
Review Before Continuing
- Does every section have a reason to exist?
- Can students tell what to do next?
- Are the three page types visibly related?
- Is the structure reusable?
- What can be removed?
Prompt 6: Build the Canvas HTML
Purpose
Now AI moves from design partner to production tool.
The major instructional and design decisions have already been made. The code should execute those decisions.
What You Need
- Your approved design system
- The approved architecture for the page you are building
- The actual content for that page
Using the approved course design system and page architecture we established in this conversation, create the Canvas-compatible HTML for this page. I am building this page type: Here is the content that belongs on this page: Follow all previously approved rules for: Information hierarchy. Visual language. Color logic and HEX codes. Page architecture. Recurring components. Accessibility. Navigation. Student calls to action. Use clean HTML appropriate for the Canvas LMS HTML editor. Use semantic heading hierarchy. Use inline styling when styling is necessary for reliable Canvas rendering. Maintain sufficient text and background contrast. Never communicate meaning through color alone. Use descriptive link text. Maintain consistent spacing and hierarchy. Keep the design readable on smaller screens. Avoid unnecessary decorative elements. Do not introduce new colors, components, terminology, fonts, or visual conventions without my approval. Do not redesign the page. Your task is to translate the design system we already approved into functioning HTML. Provide the complete HTML code in one code block so that I can copy and paste it into the Canvas HTML editor. After the code, explain the major structural components and identify anything I should specifically inspect after Canvas renders the page.
Then Test It in Canvas
Copy the HTML.
Open the Canvas HTML editor.
Paste the code.
Save the page.
Look at the rendered page.
Do not judge the code by what it looked like in the AI conversation.
Prompt 6A: Fix the Page Without Starting Over
Purpose
Something will eventually render differently than expected.
Do not throw away a mostly successful page because one element needs correction.
The Principle
Protect what works.
A revision prompt should tell AI both what to change and what it must not change.
The HTML you created is now rendering in Canvas. Most of the page should remain unchanged. Here is what is working correctly: Here is what happened when Canvas rendered the page: Here is what I expected to happen: Here is the specific element or behavior that needs to change: Revise only what is necessary to correct this problem. Preserve the approved design system, content, colors, hierarchy, spacing, components, and all elements that are already working unless changing one of them is technically necessary to solve the problem. Explain what caused the issue if you can identify it. Then provide the complete corrected HTML so I can replace the existing code in Canvas.
Prompt 7: Audit the Finished Page
Purpose
A functioning page is not necessarily an effective learning environment.
Use AI to conduct a structured critique rather than asking whether the page “looks good.”
What You Need
- The finished page
- Your approved design system
- Your page architecture
- Your course goals
Audit this Canvas page against the instructional and visual design system we established earlier in this conversation. Do not automatically praise the page. Evaluate it critically according to: 1. Instructional hierarchy. 2. Clarity. 3. Navigation. 4. Cognitive load. 5. Accessibility. 6. Color logic. 7. Visual consistency. 8. Student calls to action. 9. Mobile readability. 10. Consistency with the approved page architecture. 11. Alignment between the page design and its instructional purpose. For every problem you identify, explain: What is wrong. Why it matters for student learning or usability. How significant the problem is. What should change. Also identify elements that are working particularly well and should remain unchanged during revision. Separate necessary revisions from optional improvements. Do not rewrite the HTML until I approve the recommendations.
Review Before Revising
- Which recommendations improve learning or usability?
- Which are merely aesthetic preferences?
- Would any recommendation violate an earlier design decision?
- What should remain exactly as it is?
Scale the System
Once a page has been built, tested, revised, and approved, stop rebuilding it. Use the approved system to create the next page.
Using the approved course design system and the final approved version of the Module Overview template, create the next module. Preserve the existing: Page architecture. Visual hierarchy. Color logic. Component structure. Terminology. Navigation conventions. Accessibility standards. Do not redesign the template. Here is the content for the new module: Replace the previous module-specific content with the new content while preserving the approved system. Provide the complete Canvas-compatible HTML.
The Shift
You are no longer prompting:
Make me another Canvas page.
You are prompting:
Apply my established system to the next page.
That is how one successful page becomes a course.
When AI Gets Off Track
You do not always need a new prompt. You often need a better instruction.
If AI Makes Too Many Decisions
“Stop. You have made design or instructional decisions I did not approve. Identify those decisions and return control of them to me before continuing.”
If AI Keeps Changing Things
“Change only the element I identified. Preserve everything else unless a change is technically necessary.”
If AI Becomes Too Decorative
“Remove visual elements that do not serve an instructional, navigational, accessibility, or hierarchy function.”
If AI Gives Generic Results
“Which parts of your recommendation are generic best practices and which are specific to the course information I provided?”
If You Do Not Understand the Code
“Explain what this section of HTML does in plain language before I decide whether to use it.”
If Something Works
“Treat this version as approved. Preserve these decisions in subsequent pages unless I explicitly change them.”
The Prompting Pattern
The individual prompts matter. The pattern matters more.
Good prompting is not finding the perfect sentence. It is managing a design conversation.
Your AI Course-Building Workflow
AI helps with analysis, organization, production, coding, testing, and iteration. You decide what should be taught, how students should experience it, and whether what AI produces is good enough to use.