🤖 AI SEO optimization (2026-04-21)

[github-actions]

🤖 Automated AI SEO Content Optimization

This PR contains aggressive AI search optimization using the modernized workflow with intelligent bash pre-processing.

📊 Summary

• JSON files: 1
• Markdown files: 0
• Total changes: 3 files
• PAA data: ✅ Enabled (DataForSEO)
• Auto-fixed: None needed

🎯 AI SEO Optimizations Applied

Conversational Query Matching:

• Headings transformed to questions: "What is Mount?", "How do I escape?"
• First paragraph = direct answer box for AI snippets

Entity Recognition:

• Canonical names used consistently: Mount, Triangle Choke
• Aliases defined on first mention: "De La Riva (DLR)"
• Bold emphasis on all technique/position names

Step-by-Step Clarity:

• All techniques have 8-15 detailed steps
• Each step = one atomic action
• Sub-steps for complex movements

PAA Integration (Enabled):

• Dedicated FAQ sections at article end
• 5-10 questions from Google PAA data
• High-priority questions woven into content

[[Wikilink]] Validation:

• All links validated against content catalog
• 10-20 links per article minimum
• Prefer specific variants: [[Mount Top]] over [[Mount]]

🔧 Technical Processing

Pre-processing (Bash):

1. Generated content catalog → /tmp/preprocessing/content_catalog.txt
2. Auto-filled empty .json files (<200 bytes) with regenerate_content_json.py
3. Detected template types (SINGLE/DUAL/FAMILY positions)
4. Read templates and 5 examples per file
5. Pre-validated .json files, saved results

Claude Optimization:

• Full rewrite authority for AI SEO
• Used pre-processed context (no file discovery overhead)
• Applied entity consistency rules
• Converted headings to questions
• Added detailed execution steps

Post-processing:

• Validated all .json files
• Retry loop (max 3 attempts) for validation failures
• Auto-generated .md from validated .json
• Preserved JSON as source of truth

📝 Files Modified

$CHANGED_FILES

✅ Review Checklist

• Headings are conversational questions
• First paragraphs are direct answers (2-3 sentences)
• Entities use canonical names consistently
• All techniques have 8-15 detailed steps
• FAQ sections present (if PAA enabled)
• [[Wikilinks]] resolve correctly
• JSON validates successfully
• No dev artifacts (TODO, FIXME, etc.)
• Success rates properly ordered

---

Workflow: Content Improvement Bot (AI SEO Optimized)
Run ID: 24712971408
Timestamp: $(date -u '+%Y-%m-%dT%H:%M:%SZ')
Selection: Git creation date (oldest files)
Model: Claude Sonnet 4.5