Claude Code Subagents for Uplimit Course Creation

Two specialized Claude Code subagents to streamline the Outline → Storyboard → Build workflow for creating pedagogically-sound Uplimit courses.

🚀 Quick Start

npx @jameskruck/claude-subagents@latest --agent=all

Restart Claude Code, then use natural language:

• "Help me create a course outline for my marketing course"
• "Convert my Week 1 outline into an Uplimit storyboard"

📦 What's Included

1. Course Outline Agent

Creates robust, pedagogically-sound course outlines with:

• Course Learning Outcomes (CLOs) using Bloom's taxonomy
• Module Learning Outcomes (MLOs) supporting each CLO
• Activities aligned to Bloom's levels
• Assessments that authentically measure outcomes
• UDL principles integrated throughout

Best for: Starting new courses, revising existing outlines for QM compliance, converting course proposals into structured outlines

2. Uplimit Storyboard Agent

Converts course outlines into detailed Uplimit build specifications:

• Element selection based on Bloom's levels and UDL
• Content specifications (exact text, settings, rationale)
• Widget recommendations for interactive learning
• Accessibility design (WCAG 2.2 AA built in, not retrofitted)
• Content checklist (what needs to be created/sourced)

Best for: Converting outlines to Uplimit, designing interactive experiences, ensuring pedagogical alignment in platform builds

🎯 V3 Interactive-First Design

The Uplimit Storyboard Agent incorporates research-backed V3 Interactive-First principles:

1. No text block over 150 words (1 minute reading max)
2. Interactive widget every 2-3 elements (hands-on manipulation)
3. "Show, Don't Tell": Replace explanations with discovery experiences
4. Student agency: Let students control variables and explore outcomes
5. Progressive complexity: Start simple (lists), build to complex (simulations)

Text Reduction Target: 60-70% less text than traditional readings by converting explanations into interactive explorations.

Example Transformation

Traditional (V1):

• 3,500-word reading
• 20 minutes passive
• 5% active engagement

V3 Interactive-First:

• 1,000 words across micro-readings
• 8 min reading + 25 min widgets
• 75% active engagement ✅

📥 Installation

Option 1: Install Both Agents (Recommended)

npx @jameskruck/claude-subagents@latest --agent=all

Option 2: Install Individual Agents

# Course outline agent only
npx @jameskruck/claude-subagents@latest --agent=course-outline

# Storyboard agent only
npx @jameskruck/claude-subagents@latest --agent=uplimit-storyboard

Verify Installation

Check that agents are installed:

ls ~/.claude/agents/
# Should see: course-outline-agent.md, uplimit-storyboard-agent.md

Restart Claude Code to load the new agents.

💡 Usage Examples

Creating a Course Outline

You: "I want to create a 5-week MBA course on sports marketing.
The course should cover revenue models, sponsorship, athlete branding,
and emerging trends like betting and esports."

Agent: [Asks clarifying questions about audience, goals, key concepts]

Agent: [Creates structured outline with measurable CLOs and MLOs]

Agent: [Maps activities and assessments to each outcome]

Agent: [Provides complete outline ready for storyboarding]

Creating an Uplimit Storyboard

You: "Convert my Week 1 outline into an Uplimit storyboard. Students
need to understand 5 revenue streams in professional sports and analyze
trade-offs between them."

Agent: [Asks for CLOs, MLOs, activities, assessments]

Agent: [Recommends specific Uplimit elements with rationale]
  - Infobox for learning objectives
  - Text + Vertical List for revenue streams overview
  - iFrame widget for interactive portfolio building
  - Case study with Text Response assessment

Agent: [Provides detailed specifications for each element]
  - Exact content to type
  - Import instructions
  - Widget interaction designs
  - Rubric criteria

Agent: [Creates content checklist]
  - 3 text documents to prepare (with word counts)
  - 1 video to record (with transcript requirements)
  - 2 widgets to build (with specifications)

🎓 Complete Workflow

Step 1: Create Outline (2-3 hours)

"Help me create a course outline for [topic]"
→ Get: Complete outline with CLOs, MLOs, activities, assessments

Step 2: Validate Outline Review for:

• Measurable outcomes (single action verbs)
• Appropriate Bloom's levels (MBA = Analysis/Evaluation)
• Realistic time commitments (9-12 hours/week)
• UDL integration (multiple means of representation/engagement/expression)

Step 3: Create Storyboard (2-4 hours per module)

"Convert my Week 1 outline into an Uplimit storyboard"
→ Get: Detailed build specifications with element choices and rationale

Step 4: Prepare Content (varies) Use the content checklist from storyboard:

• Write/import text documents
• Record/source videos
• Build/embed widgets
• Create images/diagrams

Step 5: Build in Uplimit (4-6 hours per module) Follow your storyboard exactly:

• Create structure (units, modules)
• Import text content
• Add visual elements
• Embed interactive elements
• Configure assessments

Step 6: Test & Launch (1-2 hours)

• Walk through as student
• Test all interactions
• Verify accessibility
• Get peer review

⚡ Benefits vs. Manual Process

Without Agents (Traditional)

• ❌ 8-12 hours per module
• ❌ Frequent rework due to misalignment
• ❌ Vague outcomes that don't measure well
• ❌ Accessibility retrofitted (often incomplete)
• ❌ Inconsistent UDL application
• ❌ Trial-and-error element selection

With Agents (Systematic)

• ✅ 4-6 hours per module (40% faster)
• ✅ Alignment verified before building
• ✅ Measurable outcomes from the start
• ✅ Accessibility designed in
• ✅ Systematic UDL integration
• ✅ Pedagogically-justified element choices

Time Saved per 5-Module Course: 20-30 hours

Quality Improvements:

• Higher QM compliance (measurable outcomes, aligned assessments)
• Better UDL implementation (intentional, not accidental)
• Stronger accessibility (WCAG 2.2 AA by design)
• Clearer pedagogical rationale (defendable choices)

📚 Educational Frameworks

Both agents apply:

• Quality Matters (QM) Standards: Measurable outcomes, aligned assessments, clear objectives
• Universal Design for Learning (UDL): Multiple means of representation, engagement, and expression
• Bloom's Taxonomy: Appropriate cognitive levels for target audience
• WCAG 2.2 AA: Accessibility by design, not retrofit

🔧 Troubleshooting

Agent Not Responding

1. Verify installation: ls ~/.claude/agents/
2. Restart Claude Code completely
3. Reinstall: npx @jameskruck/claude-subagents@latest --agent=all

Wrong Agent Activated

Be explicit: "Use the Uplimit Storyboard Agent to convert this outline"

Agent Missing Context

Repeat key details (CLOs, MLOs, activities) explicitly in your message

📖 Documentation

For complete documentation, see:

• Course Outline Agent - Full agent prompt
• Uplimit Storyboard Agent - Full agent prompt
• Installation Guide - Detailed setup instructions

🤝 Support

Questions? Ask Claude Code (it has all training materials)

Issues? Open issue at: https://github.com/jameskruck/claude-subagents/issues

Want to Contribute? Share your storyboards and patterns!

📄 License

MIT © James Kruck

Streamlining pedagogically-sound Uplimit course creation for Learning Experience Designers