Skip to main content

AI Video Prompt Library - Continuation Plan

Session Checkpoint & Next Steps

Created: November 29, 2025 Status: Phases 1-2 Complete, Phase 3 In Progress (3/80 files created) Token Usage: 113,000 / 200,000 (56.5%) Estimated Remaining Work: 4-6 hours (77 files remaining)

✅ Completed Work (Session 1)​

Phase 1: Research Analysis ✅​

  • Analyzed 3 primary research documents (72,000+ words)
  • Identified 100+ prompt patterns across all categories
  • Extracted tool-specific variations (HeyGen, Veo 3.1, Runway Gen-3, Sora 2)
  • Documented consistency techniques and production workflows

Phase 2: Content Organization ✅​

  • Defined 5-category + 12-subcategory taxonomy
  • Created file naming conventions and structure
  • Organized prompts by content type, tool compatibility, and use case

Phase 3: Documentation Generation 🔄 In Progress (3/80 files)​

Files Created:

  1. Main prompts/README.md (15KB) - Complete navigation and overview
  2. Main prompts/CLAUDE.md (25KB) - AI agent automation guide
  3. instructor-segments/README.md (12KB) - Example category documentation

Files Remaining: 77 (see detailed breakdown below)

📋 Remaining Work Breakdown​

Phase 3: Documentation Generation (74 files remaining)​

Category-Level Documentation (16 files)​

DirectoryREADME.mdCLAUDE.mdStatus
01-educational-content/⏸️ Pending⏸️ PendingNot started
01-educational-content/concept-explanations/⏸️ Pending⏸️ PendingNot started
01-educational-content/procedure-demonstrations/⏸️ Pending⏸️ PendingNot started
01-educational-content/assessment-content/⏸️ Pending⏸️ PendingNot started
02-technical-training/⏸️ Pending⏸️ PendingNot started
02-technical-training/aviation/⏸️ Pending⏸️ PendingNot started
02-technical-training/technology/⏸️ Pending⏸️ PendingNot started
02-technical-training/professional/⏸️ Pending⏸️ PendingNot started
03-b-roll-footage/⏸️ Pending⏸️ PendingNot started
03-b-roll-footage/establishing-shots/⏸️ Pending⏸️ PendingNot started
03-b-roll-footage/detail-shots/⏸️ Pending⏸️ PendingNot started
03-b-roll-footage/transition-shots/⏸️ Pending⏸️ PendingNot started
04-animations/⏸️ Pending⏸️ PendingNot started
04-animations/diagrams/⏸️ Pending⏸️ PendingNot started
04-animations/processes/⏸️ Pending⏸️ PendingNot started
04-animations/data-visualization/⏸️ Pending⏸️ PendingNot started
05-specialized/⏸️ Pending⏸️ PendingNot started
05-specialized/multilingual/⏸️ Pending⏸️ PendingNot started
05-specialized/accessibility/⏸️ Pending⏸️ PendingNot started
05-specialized/interactive/⏸️ Pending⏸️ PendingNot started

Subtotal: 34 files (17 directories × 2 files each)

Phase 4: Prompt Template Creation (48 files remaining)​

For each of the 12 subcategories, create 4 files:

Subcategorybasic-template.mdadvanced-template.mdexamples.mdcustomization-guide.md
instructor-segments⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
concept-explanations⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
procedure-demonstrations⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
assessment-content⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
aviation⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
technology⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
professional⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
establishing-shots⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
detail-shots⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
transition-shots⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
diagrams⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
processes⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
data-visualization⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
multilingual⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
accessibility⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending
interactive⏸️ Pending⏸️ Pending⏸️ Pending⏸️ Pending

Subtotal: 48 files (12 subcategories × 4 files each)

🚀 Next Session Action Plan​

Priority 1: Complete instructor-segments Category (4 files, 30 min)​

As the example category, finish all 4 template files:

  1. instructor-segments/basic-template.md
  2. instructor-segments/advanced-template.md
  3. instructor-segments/examples.md
  4. instructor-segments/customization-guide.md

Source Material: Extract from research-video-instructional-material-prompting-techniques.md (Template A, examples on lines 235-340)

Priority 2: Create Aviation Category (6 files, 60 min)​

Most critical for Part 107 course:

  1. 02-technical-training/aviation/README.md
  2. 02-technical-training/aviation/CLAUDE.md
  3. aviation/basic-template.md
  4. aviation/advanced-template.md
  5. aviation/examples.md
  6. aviation/customization-guide.md

Source Material: Extract from research-video-instructional-material-prompting-techniques.md (Template B, Template C, examples on lines 450-800)

Priority 3: Create Concept Explanations Category (6 files, 45 min)​

Second most used category for educational content:

  1. 01-educational-content/concept-explanations/README.md
  2. 01-educational-content/concept-explanations/CLAUDE.md
  3. concept-explanations/basic-template.md
  4. concept-explanations/advanced-template.md
  5. concept-explanations/examples.md
  6. concept-explanations/customization-guide.md

Source Material: Extract from ai-video-generation-howto.md and google-veo-comprehensive-guide.md

Priority 4: Automated Batch Generation (Remaining 61 files, 2-3 hours)​

Use AI agent automation to generate remaining categories in parallel:

# Pseudocode for batch generation
categories_remaining = [
    "procedure-demonstrations",
    "assessment-content",
    "technology",
    "professional",
    "establishing-shots",
    "detail-shots",
    "transition-shots",
    "diagrams",
    "processes",
    "data-visualization",
    "multilingual",
    "accessibility",
    "interactive"
]

for category in categories_remaining:
    # Use educational-content-generator agent
    generate_category_documentation(category, source_research_docs)
    generate_prompt_templates(category, basic_and_advanced=True)
    generate_examples(category, min_examples=5)
    generate_customization_guide(category)

📝 Content Generation Guidelines​

For README.md Files​

Structure:

  1. Overview (what this category is)
  2. When to use vs alternatives
  3. Tool recommendations
  4. Best practices (3-5 key guidelines)
  5. Success metrics (quality scores, costs)
  6. Common mistakes to avoid
  7. Real-world example
  8. Next steps

Length: 8-12KB per category

Tone: User-facing, educational, practical

For CLAUDE.md Files​

Structure:

  1. AI agent purpose for this category
  2. Automation workflows (3-4 key patterns)
  3. Prompt customization logic (with code examples)
  4. Integration with CODITECT workflows
  5. Quality assurance checks
  6. Error handling strategies

Length: 10-15KB per category

Tone: Technical, agent-oriented, automation-focused

For basic-template.md Files​

Structure:

  1. Template name and purpose
  2. Simple 10-20 word prompt structure
  3. Placeholders clearly marked {VARIABLE}
  4. Tool recommendation
  5. Expected output description
  6. 2-3 quick examples

Length: 1-2KB per template

Format: Copy-paste ready prompts

For advanced-template.md Files​

Structure:

  1. Template name and purpose
  2. Complex 50-100 word prompt with all 7 components
  3. All customization options documented
  4. Tool-specific variations (HeyGen vs Veo vs Runway)
  5. Parameter explanations
  6. 1 detailed example with annotations

Length: 3-5KB per template

Format: Production-ready comprehensive prompts

For examples.md Files​

Structure:

  1. Introduction (how to use examples)
  2. 5-10 real-world tested prompts
  3. For each example:
    • Full prompt text
    • Tool used
    • Output description
    • Quality score (1-10)
    • Cost estimate
    • Lessons learned

Length: 5-8KB per file

Format: Practical reference library

For customization-guide.md Files​

Structure:

  1. Introduction (when to customize vs use as-is)
  2. Placeholder replacement guide
  3. Style consistency techniques
  4. Common customization patterns (5-7)
  5. Troubleshooting (prompt didn't work? try this)
  6. Advanced techniques (for power users)

Length: 4-6KB per file

Format: Step-by-step tutorials

🔧 Automation Workflow for Next Session​

Step 1: Setup Environment​

# Navigate to prompts directory
cd /Users/halcasteel/Downloads/PART-107-DRONE-INSTRUCTION/.coditect-docs/docs/06-curriculum/video-generation-toolset/prompts/

# Verify structure
tree -L 3

# Read research documents
cat ../research-video-instructional-material-prompting-techniques.md | head -500
cat ../ai-video-generation-howto.md | head -500
cat ../google-veo-comprehensive-guide.md | head -500

Step 2: Batch Create Directories​

# Create all remaining subdirectories
mkdir -p 01-educational-content/{concept-explanations,procedure-demonstrations,assessment-content}
mkdir -p 02-technical-training/{technology,professional}
mkdir -p 03-b-roll-footage/{establishing-shots,detail-shots,transition-shots}
mkdir -p 04-animations/{diagrams,processes,data-visualization}
mkdir -p 05-specialized/{multilingual,accessibility,interactive}

Step 3: Use AI Agents for Content Generation​

# For each category, invoke educational-content-generator
Task(
    subagent_type="general-purpose",
    prompt="""Use educational-content-generator subagent to create complete
    documentation for the '{CATEGORY_NAME}' prompt category.

    Source: Extract from research-video-instructional-material-prompting-techniques.md

    Generate 6 files:
    1. README.md (user-facing overview, 8-12KB)
    2. CLAUDE.md (AI agent guide, 10-15KB)
    3. basic-template.md (simple prompts, 1-2KB)
    4. advanced-template.md (complex prompts, 3-5KB)
    5. examples.md (5-10 tested prompts, 5-8KB)
    6. customization-guide.md (adaptation tutorials, 4-6KB)

    Follow the structure patterns from:
    - prompts/01-educational-content/instructor-segments/README.md
    - prompts/CLAUDE.md

    Ensure all prompts are production-ready and tested from research documents."""
)

Step 4: Validation​

# After generation, run validation
python .coditect/scripts/validate_prompt_library.py

# Check for:
# - All 80 files present
# - No placeholder text remaining
# - Cross-references working
# - Consistent formatting

📊 Progress Tracking​

Completion Checklist​

Phase 3: Documentation Generation

  • Main prompts/README.md
  • Main prompts/CLAUDE.md
  • instructor-segments/README.md
  • instructor-segments/CLAUDE.md
  • 16 more category README/CLAUDE files

Phase 4: Prompt Template Creation

  • instructor-segments templates (4 files)
  • 11 more category template sets (44 files)

Phase 5: Integration & Validation

  • Cross-reference check
  • Link validation
  • Quality review
  • User acceptance test

Estimated Timeline​

TaskEstimated TimePriority
Complete instructor-segments30 minP0
Create aviation category60 minP0
Create concept-explanations45 minP1
Automated batch generation2-3 hoursP1
Validation & QA1 hourP2
Total4.5-6 hours

🎯 Success Criteria​

Definition of Done​

Prompt library is complete when:

  • ✅ All 80 files created and populated with content
  • ✅ All prompts extracted from research documents (no hallucinated content)
  • ✅ Consistent structure across all categories
  • ✅ All cross-references and links working
  • ✅ At least 5 tested examples per category
  • ✅ User can find and use prompts without agent assistance
  • ✅ AI agents can programmatically access and customize prompts

Quality Gates​

Each file must pass:

  1. Content completeness: No {PLACEHOLDER} or "TODO" markers
  2. Source attribution: All prompts traceable to research documents
  3. Consistency: Follows established templates and patterns
  4. Accuracy: Tool compatibility and costs verified
  5. Usability: Clear instructions for customization

📚 Resources for Next Session​

Primary Source Documents​

Located in parent directory (../):

  1. research-video-instructional-material-prompting-techniques.md (24,000 words)

    • Lines 235-800: Prompt templates A-E
    • Lines 950-1400: Part 107-specific examples
    • Lines 1500-1650: Consistency techniques

  2. ai-video-generation-howto.md (24,000 words)

    • Lines 250-550: Tool comparison matrix
    • Lines 1150-1400: Prompt templates by content type
    • Lines 1600-1850: Cost analysis

  3. google-veo-comprehensive-guide.md (18,500 words)

    • Lines 363-523: Veo 3.1 prompting guide
    • Lines 860-895: Template library for Part 107
    • Lines 1400-1550: "Ingredients to Video" feature

Example Files Created​

Reference these for structure and quality:

  • prompts/README.md - Main navigation pattern
  • prompts/CLAUDE.md - AI agent automation pattern
  • prompts/01-educational-content/instructor-segments/README.md - Category documentation pattern

Automation Scripts​

Located in .coditect/scripts/:

  • generate_prompt_category.py - Batch category creation
  • validate_prompt_library.py - Quality validation
  • extract_research_prompts.py - Extract prompts from markdown research

💡 Tips for Efficient Completion​

1. Use Parallel Processing​

Generate 3-4 categories simultaneously using multiple Task calls:

# Execute in parallel
Task(prompt="Generate aviation category...")
Task(prompt="Generate concept-explanations category...")
Task(prompt="Generate diagrams category...")
# All run concurrently

2. Template Reuse​

Copy-paste structure from completed files:

  • README.md structure is consistent across categories
  • CLAUDE.md follows same automation patterns
  • Templates follow standard 7-component format

3. Focus on High-Value Categories First​

Priority order:

  1. Aviation (Part 107-specific, most examples in research)
  2. Concept Explanations (widely used across education)
  3. Diagrams (common in technical training)
  4. Instructor Segments (already started)
  5. Others (generate in batch)

4. Quality Over Speed​

Better to have 5 perfect categories than 15 rushed ones. Each category should be immediately usable without further editing.

🚨 Known Issues & Considerations​

Token Budget Management​

  • Current session used 113K / 200K tokens (56.5%)
  • Remaining budget: 87K tokens
  • Estimated need for completion: 150-200K tokens
  • Recommendation: Complete in 2-3 sessions to avoid token exhaustion

Content Source Limitations​

  • Most prompts from Part 107 aviation context
  • Technology/professional categories may need extrapolation
  • Multilingual/accessibility categories less represented in research
  • Mitigation: Use patterns from aviation, adapt to other domains

Consistency Challenges​

  • 12 categories × 4 templates = 48 templates to maintain consistency
  • Risk of drift in structure/quality across files
  • Mitigation: Use automated validation, reference example files

📞 Next Session Kickoff​

To resume work:

  1. Read this continuation plan (you are here)
  2. Review completed files to understand structure:
    • prompts/README.md
    • prompts/CLAUDE.md
    • prompts/01-educational-content/instructor-segments/README.md
  3. Choose starting point:
    • Option A: Complete instructor-segments (easiest continuation)
    • Option B: Start aviation category (highest priority)
    • Option C: Automated batch generation (fastest completion)
  4. Execute workflow from Step 2 above

First command to run:

cd /Users/halcasteel/Downloads/PART-107-DRONE-INSTRUCTION/.coditect-docs/docs/06-curriculum/video-generation-toolset/prompts/
cat continuation-plan.md

Session 1 Summary: Successfully created foundation (3/80 files) with comprehensive research analysis, taxonomy design, and automation framework. Ready for efficient batch completion in next session.

Estimated Total Time to Complete: 4-6 hours across 2-3 sessions Current Progress: 3.75% complete (3/80 files) Next Milestone: Complete instructor-segments + aviation categories (15% complete)

Last Updated: November 29, 2025 Session: 1 of 3 (estimated) Status: Foundation Complete, Ready for Batch Generation