Fitness Influencer Operations

⚡ Quick Reference (AI: Read This First)

CAPABILITIES: Video editing (jump cuts) • Educational graphics • Email digest • Revenue analytics • AI images (Grok) • Video ads (Shotstack) • Workout plans • Nutrition guides • Video Blueprints • Content Calendar • Comment Categorizer • Cross-Platform Optimizer

WHEN TO USE: User requests content creation, video editing, email management, or analytics for fitness influencer workflows

EXECUTION: Uses deterministic Python scripts in execution/ directory. Reference directive directives/fitness_influencer_operations.md for complete details.

🎯 Decision Tree

User Request → Capability Mapping:

├─ "edit video" / "jump cuts" / "remove silence"
│  └─ USE: execution/video_jumpcut.py
│     INPUT: Video file, silence threshold
│     OUTPUT: Edited video, stats
│
├─ "create graphic" / "fitness tip card" / "educational content"
│  └─ USE: execution/educational_graphics.py
│     INPUT: Title, key points, platform
│     OUTPUT: Branded graphic (1080x1080, 1080x1920, or 1280x720)
│
├─ "email" / "inbox" / "summarize emails"
│  └─ USE: execution/gmail_monitor.py
│     INPUT: Time period (hours back)
│     OUTPUT: Categorized email digest
│
├─ "revenue" / "analytics" / "expenses" / "profit"
│  └─ USE: execution/revenue_analytics.py
│     INPUT: Google Sheet ID, month
│     OUTPUT: Revenue report with trends
│
├─ "generate image" / "AI image" / "fitness photo"
│  └─ USE: execution/grok_image_gen.py
│     INPUT: Prompt, count
│     OUTPUT: Image URLs ($0.07 each)
│
├─ "video ad" / "create ad" / "social video" / "advertisement"
│  └─ USE: execution/video_ads.py (orchestrates Grok + Shotstack)
│     PROCESS: Generate 4 AI images → Create video with transitions
│     OUTPUT: Video URL ($0.34 = 4×$0.07 images + $0.06 video)
│
├─ "workout plan" / "training program" / "exercise routine"
│  └─ USE: execution/workout_plan_generator.py
│     INPUT: Goal, experience level, days/week, equipment
│     OUTPUT: Structured workout plan (markdown + JSON)
│
├─ "nutrition plan" / "meal plan" / "macros" / "diet guide"
│  └─ USE: execution/nutrition_guide_generator.py
│     INPUT: Weight, activity level, goal, dietary preference
│     OUTPUT: Personalized nutrition guide (markdown + JSON)
│
├─ "video blueprint" / "video template" / "plan my video"
│  └─ USE: execution/video_template_framework.py
│     INPUT: Topic, style, duration, platform
│     OUTPUT: Segment-by-segment script with visual hints (HTML timeline)
│
├─ "content calendar" / "posting schedule" / "plan content"
│  └─ USE: MCP tool: generate_content_calendar
│     INPUT: Days, posts/day, platforms, content focus, rest days
│     OUTPUT: 30-day balanced content plan with burnout prevention
│
├─ "categorize comments" / "sort DMs" / "manage engagement"
│  └─ USE: MCP tool: categorize_comments
│     INPUT: Array of comment strings
│     OUTPUT: Categorized comments with auto-reply suggestions
│
└─ "optimize for platforms" / "cross-platform" / "repurpose content"
   └─ USE: MCP tool: optimize_for_platforms
      INPUT: Content type, caption, video duration, hashtags
      OUTPUT: Platform-specific optimizations for 9 platforms

📋 Script Reference

🔧 Environment Requirements

Required in .env:

# AI Image Generation
XAI_API_KEY=xxx                    # Grok/xAI for images

# Video Generation
SHOTSTACK_API_KEY=xxx              # Shotstack for video ads
SHOTSTACK_ENV=stage                # or 'v1' for production

# Google APIs (optional)
GOOGLE_CLIENT_ID=xxx               # For Gmail, Calendar, Sheets
GOOGLE_CLIENT_SECRET=xxx
GOOGLE_REFRESH_TOKEN=xxx

# Canva API (optional)
CANVA_API_KEY=xxx                  # For advanced designs

🔍 Unknown Use Case Handler

If user request doesn't match the Decision Tree above:

1. Analyze Request Components

• Extract: action verb, target object, desired outcome
• Example: "Create workout plan PDF" → action=create, object=PDF, outcome=workout plan

2. Check Capability Proximity

Ask yourself:

• Can existing scripts be combined to achieve this?
• Is this a variation of a known capability?
• Does this require new development?

3. Response Options

Option A: Combinable (Best Case)

"I can accomplish this by combining existing tools:
1. Use educational_graphics.py to create workout content
2. Use markdown_to_pdf.py to format as PDF
Would you like me to proceed with this approach?"

Option B: Close Match

"I don't have an exact tool for this, but I can do [similar capability].
For example, I can create workout graphics but not full PDF plans yet.
Would that work for your needs?"

Option C: New Capability Needed

"This requires new development. I'll log it as a capability gap.
Based on frequency data, this has been requested [X] times.
I recommend creating a dedicated tool for this workflow."

4. Learning Protocol

For every unhandled request:

1. Log to USE_CASES.json → unhandled_requests[]
2. Analyze if it fits existing patterns or needs new capability
3. Update frequency counters for capability gaps
4. Suggest directive updates for recurring patterns
5. Document what was learned in learning_log[]

5. Auto-Documentation Update

When a new pattern emerges (frequency > 3):

1. Add to known_use_cases[] in USE_CASES.json
2. Update Decision Tree with new routing
3. Add Usage Example showing the workflow
4. Document in learning_log[] what made this work

This creates a self-improving system - the more it's used, the smarter it gets!

💡 Usage Examples

Example 1: Edit Video with Jump Cuts

User: "Edit this raw video and remove the silence"

AI Decision:
  ✓ Matches: video editing capability
  ✓ Execute: video_jumpcut.py
  ✓ Default threshold: -40dB

Command:
  python execution/video_jumpcut.py \
    --input uploaded_video.mp4 \
    --output edited_video.mp4 \
    --silence-thresh -40

Output:
  - Edited video with jump cuts
  - Stats: original 15:30 → 8:45 (43% reduction, 47 cuts)

Example 2: Create Fitness Tip Graphic

User: "Create an Instagram post about staying lean without tracking macros"

AI Decision:
  ✓ Matches: educational graphics capability
  ✓ Extract: title + key points from request
  ✓ Execute: educational_graphics.py

Command:
  python execution/educational_graphics.py \
    --title "Staying Lean Without Tracking Macros" \
    --points "Focus on whole foods,Eat protein with every meal,Stay hydrated" \
    --platform instagram_post

Output:
  - 1080x1080 branded graphic with Marceau Solutions branding

Example 3: Generate Video Ad

User: "Create a 15-second video ad for @boabfit"

AI Decision:
  ✓ Matches: video ad creation capability
  ✓ Pipeline: Grok images → Shotstack video
  ✓ Execute: Two-step process

Commands:
  # Step 1: Generate 4 AI images
  python execution/grok_image_gen.py \
    --prompt "Athletic fitness influencer in modern gym" \
    --count 4

  # Step 2: Create video from images
  python execution/shotstack_api.py create-fitness-ad \
    --images "url1,url2,url3,url4" \
    --headline "Transform Your Body" \
    --cta "Follow @boabfit"

Output:
  - 14-second vertical video (9:16)
  - Cost: $0.34 ($0.28 images + $0.06 video)
  - Ready for Instagram/TikTok

📚 Complete Documentation

For comprehensive details, see:

• Directive: directives/fitness_influencer_operations.md (600+ lines)
• Tech Evaluation: .claude/FITNESS_INFLUENCER_TECH_EVALUATION.md
• Deployment Plan: .claude/FITNESS_INFLUENCER_DEPLOYMENT_PLAN.md
• Knowledge Base: .claude/KNOWLEDGE_BASE.md

🎉 Recent Successes

2026-01-06: First successful video ad created

• 4 Grok-generated images → Shotstack video
• @boabfit promotional ad (14 seconds)
• Total cost: $0.34
• Video URL: https://shotstack-api-stage-output.s3-ap-southeast-2.amazonaws.com/...

2026-01-05: Backend deployed to Railway

• Video processing with static file serving
• All execution scripts tested and working
• API endpoints functional

⚙️ Technical Notes

Deployment: Deployed from DOE development environment on 2026-01-05

Status: Production-ready with following capabilities:

• ✅ Video editing (FFmpeg + MoviePy)
• ✅ Graphics generation (Pillow)
• ✅ AI images (Grok/xAI API)
• ✅ Video ads (Shotstack API)
• ✅ Email monitoring (Gmail API)
• ✅ Revenue analytics (Google Sheets API)
• ⏳ Calendar reminders (Google Calendar API - setup pending)
• ⏳ Canva integration (API available - implementation pending)

Intermediate Files: All temporary files stored in .tmp/ (not committed)

Error Handling: Scripts include comprehensive error handling and logging

Testing: All core features tested and verified working