article-seo-leo-enhancer

name	article-seo-leo-enhancer
description	Analyze and enhance markdown articles with SEO/LEO (LLM Engine Optimization) frontmatter fields. Adds AI-Summary, Key-Takeaways, Article-Type, Expertise-Level, FAQ, and HowTo-Steps based on article content. Automatically invokes article-post-edit-actions skill after completion.

Article SEO/LEO Enhancer

Analyze article content and add appropriate SEO/LEO frontmatter fields to improve search engine visibility and AI/LLM discoverability.

When to Use

• User asks to optimize an article for SEO
• User wants to add AI-friendly metadata to an article
• User asks to enhance an article with structured data
• Preparing an article for publication with full SEO/LEO optimization
• User mentions adding FAQ schema, HowTo schema, or AI summary

Inputs

• file_path: Path to the markdown article file
• fields (optional): Specific fields to add (default: analyze and add all applicable)

Supported SEO/LEO Fields

Field	Purpose	When to Add
Article-Type	Schema.org type selection	Always
Expertise-Level	AI context for audience	Always
AI-Summary	2-sentence summary for LLM citation	Always
Key-Takeaways	Structured insights (3-5 points)	Educational/technical content
Topics	Semantic topics beyond tags	When tags are generic
FAQ	FAQPage schema data	Articles with Q&A content
HowTo-Steps	HowTo schema data	Tutorials/guides with steps

Process

Step 1: Read and Analyze Article

1. Read the markdown file
2. Parse front matter (YAML between --- markers)
3. Identify existing SEO/LEO fields (skip if already present)
4. Analyze content structure:
   • Headings hierarchy
   • Presence of step-by-step instructions
   • Q&A patterns
   • Code examples
   • Technical depth

Step 2: Determine Article-Type

Analyze content to select appropriate type:

Content Pattern	Article-Type
Step-by-step instructions, "How to" in title	howto
Multiple Q&A sections, FAQ heading	faq
Deep technical explanation, code-heavy	tutorial
General informational content	article

Check for:

• Title contains "How to", "Guide to", "Tutorial"
• Numbered steps or ordered lists
• Headings that are questions
• <!-- faq-start --> or <!-- howto-start --> markers

Step 3: Determine Expertise-Level

Analyze based on:

Indicator	Level
Basic concepts, introductory language, no prerequisites	beginner
Some assumed knowledge, moderate depth	intermediate
Complex concepts, advanced techniques, expert audience	advanced
Cutting-edge, research-level, assumes mastery	expert

Check for:

• Prerequisite mentions
• Complexity of code examples
• Technical vocabulary density
• Assumed knowledge references

Step 4: Generate AI-Summary

Create a 2-sentence summary optimized for AI citation:

Format:

AI-Summary: >
  [Sentence 1: What the article covers/teaches].
  [Sentence 2: Key insight, main takeaway, or unique value].

Guidelines:

• First sentence: Factual statement of article scope
• Second sentence: Key insight or actionable takeaway
• Make it standalone (reader should understand value without reading article)
• Avoid marketing language, be factual
• Max 300 characters total

Example:

AI-Summary: >
  This article explains techniques to boost RAG pipeline performance
  in production environments. Key optimizations include hybrid search,
  re-ranking, and addressing the lost-in-the-middle problem.

Step 5: Extract Key-Takeaways

Identify 3-5 main points from the article:

Sources for takeaways:

• Main section headings (## level)
• Explicit conclusions or recommendations
• TL;DR section if present
• Summary paragraph

Format:

Key-Takeaways:
  - [Actionable insight 1]
  - [Actionable insight 2]
  - [Actionable insight 3]

Guidelines:

• Start with action verbs when possible
• Be specific, not generic
• Each point should stand alone
• Focus on what reader will learn/gain

Step 6: Extract FAQ (if applicable)

Trigger conditions:

• Article has Q&A sections
• Headings are questions (contain "?")
• <!-- faq-start --> marker present
• Article-Type is faq

Extract from:

1. <!-- faq-start --> ... <!-- faq-end --> markers (priority)
2. Headings that are questions + following paragraphs
3. Explicit Q&A formatted sections

Format:

FAQ:
  - question: What is X?
    answer: >
      X is [concise answer, 1-3 sentences].
  - question: How do I Y?
    answer: >
      You can Y by [actionable steps].

Guidelines:

• Keep answers concise (2-4 sentences ideal for featured snippets)
• Answers should be self-contained
• Extract 3-7 most relevant Q&A pairs
• Escape special characters in YAML

Step 7: Extract HowTo-Steps (if applicable)

Trigger conditions:

• Article-Type is howto
• Contains numbered steps or ordered lists
• <!-- howto-start --> marker present
• Title contains "How to"

Extract from:

1. <!-- howto-start --> ... <!-- howto-end --> markers (priority)
2. Numbered headings (### Step 1:, ### 1., etc.)
3. Ordered lists with substantial content

Format:

HowTo-Steps:
  - name: Step title without number prefix
    text: Brief description of what to do (1-2 sentences)
  - name: Next step
    text: Description

Guidelines:

• Remove "Step N:" prefixes from names
• Keep text concise (max 200 chars)
• Include 3-10 steps
• Steps should be actionable

Step 8: Add Topics (if beneficial)

When to add:

• Tags are too generic (e.g., just "python", "ml")
• Article covers specific subtopics not in tags
• Semantic topics would help AI understanding

Format:

Topics:
  - specific-topic-1
  - specific-topic-2
  - specific-topic-3

Guidelines:

• Use kebab-case
• Be more specific than tags
• 3-5 topics maximum
• Think: "What would someone search to find this?"

Step 9: Update Front Matter

Insert new fields in the front matter in this order (after existing fields):

---
Title: ...
Date: ...
Modified: ...
tags: ...
Category: ...
Image: ...
Summary: ...
Status: published

# SEO/LEO Enhancement
Article-Type: howto
Expertise-Level: intermediate
AI-Summary: >
  Two sentence summary here.
Key-Takeaways:
  - Point 1
  - Point 2
  - Point 3
Topics:
  - topic-1
  - topic-2
FAQ:
  - question: Q1?
    answer: A1
HowTo-Steps:
  - name: Step name
    text: Step description
---

Rules:

• Add comment # SEO/LEO Enhancement before new fields
• Only add fields that are applicable
• Don't duplicate existing fields
• Preserve all original front matter

Step 10: Invoke Post-Edit Actions

After completing SEO/LEO enhancement, invoke the article-post-edit-actions skill with:

• file_path: Same article file
• edit_descriptions: List of changes made, e.g.:
  • "Added AI-Summary for LLM optimization"
  • "Added Key-Takeaways (5 points)"
  • "Added FAQ schema with 4 Q&A pairs"
  • "Added HowTo-Steps (6 steps)"
  • "Set Article-Type to howto"
  • "Set Expertise-Level to intermediate"

Example Transformation

Before:

---
Title: Techniques to Boost RAG Performance in Production
Date: 2023-11-01
tags:
  - machine-learning
  - rag
  - llm
Category: Generative AI
Image: /images/head/boosting_RAG.jpg
Summary: This article discusses several advanced techniques...
Status: published
---

After:

---
Title: Techniques to Boost RAG Performance in Production
Date: 2023-11-01
Modified: 2026-02-07
tags:
  - machine-learning
  - rag
  - llm
Category: Generative AI
Image: /images/head/boosting_RAG.jpg
Summary: This article discusses several advanced techniques...
Status: published

# SEO/LEO Enhancement
Article-Type: article
Expertise-Level: intermediate
AI-Summary: >
  This article covers advanced techniques to optimize RAG pipeline 
  performance in production, including hybrid search, re-ranking, 
  and chunking strategies. Key insight: addressing the lost-in-the-middle 
  problem can significantly improve LLM output quality.
Key-Takeaways:
  - Combine semantic and keyword search with hybrid search for better retrieval
  - Use re-ranking to diversify retrieved snippets
  - Fine-tune embedding models for domain-specific improvements
  - Address lost-in-the-middle by reordering context snippets
  - Implement query transformations for complex queries
Topics:
  - retrieval-augmented-generation
  - vector-search-optimization
  - llm-context-management
  - embedding-fine-tuning
---

Output

• Modified article with SEO/LEO frontmatter fields
• Summary of added fields and their values
• Post-edit actions invoked for metadata updates

Success Criteria

• Article-Type correctly identified based on content
• Expertise-Level appropriate for content complexity
• AI-Summary is 2 sentences, factual, standalone
• Key-Takeaways are specific and actionable (3-5 points)
• FAQ extracted only when Q&A content exists
• HowTo-Steps extracted only for tutorial/guide content
• YAML formatting is valid (proper indentation, escaping)
• Original front matter preserved
• Post-edit actions skill invoked with change descriptions
• Modified date updated

Edge Cases

Existing SEO/LEO Fields

• Skip fields that already exist
• Report which fields were skipped
• Only add missing fields

No Clear Article-Type

• Default to article
• Note in output that type was defaulted

Short Articles

• Still add Article-Type and Expertise-Level
• AI-Summary may be very concise
• Key-Takeaways may have only 2-3 points

Mixed Content (both FAQ and HowTo)

• Can add both FAQ and HowTo-Steps
• Article-Type should be primary focus (usually howto)

Special Characters in Content

• Escape quotes in YAML: " → \"
• Use > for multi-line strings
• Avoid : at start of lines in values

Field Reference Quick Guide

# Always add:
Article-Type: article | howto | faq | tutorial
Expertise-Level: beginner | intermediate | advanced | expert
AI-Summary: >
  Two sentences. First states coverage, second gives key insight.

# Add for educational content:
Key-Takeaways:
  - Actionable point 1
  - Actionable point 2
  - Actionable point 3

# Add when tags are generic:
Topics:
  - specific-topic-1
  - specific-topic-2

# Add when Q&A content exists:
FAQ:
  - question: Question text?
    answer: Concise answer text.

# Add for tutorials/guides:
HowTo-Steps:
  - name: Step title
    text: Step description

Notes

• This skill chains to article-post-edit-actions on completion
• Focus on quality over quantity - only add applicable fields
• AI-Summary is the most impactful field for LLM discoverability
• Test generated YAML with a YAML validator if unsure
• Reference: See pelican-themes/Flex/docs/SEO_LEO_CHEATSHEET.md for full field documentation