dojo.md 0.1.0

package/courses/blog-post-seo-writing/scenarios/level-3/multilingual-seo-content.yaml

@@ -0,0 +1,78 @@
+meta:
+  id: multilingual-seo-content
+  level: 3
+  course: blog-post-seo-writing
+  type: output
+  description: "Create SEO blog content for multiple languages and regions"
+  tags: [seo, multilingual, international, hreflang, localization]
+
+state: {}
+
+trigger: |
+  International SEO requires more than translation — it requires
+  localization of keywords, search intent, and content structure for
+  each target market. A direct translation often targets the wrong
+  keywords entirely.
+
+  You're expanding a fintech company's blog from English (US) to:
+  - Spanish (Spain + Latin America)
+  - German (Germany)
+  - Portuguese (Brazil)
+
+  Design the multilingual SEO content strategy:
+
+  1. Keyword localization (not just translation):
+     Take this English blog post: "How to Build an Emergency Fund"
+     (target keyword: "emergency fund", 40,000 monthly searches US)
+
+     For each language:
+     - Research the correct local keyword (not direct translation)
+     - Explain why direct translation often targets wrong keywords
+     - Local search volume comparison
+     - Cultural context differences that affect content angle
+     - Example: In Germany, "Notgroschen" vs "Notfallfonds" — which
+       people actually search for and why
+
+  2. Content adaptation strategy:
+     - What content can be translated with localization
+     - What content needs to be rewritten from scratch
+     - Cultural differences that change the content angle
+     - Local regulations that affect financial content (YMYL compliance per region)
+     - Local examples, currencies, and financial systems
+
+  3. Technical implementation:
+     - Hreflang tag implementation (show exact code)
+     - URL structure decision: subdirectories (/es/) vs subdomains (es.site.com)
+       vs ccTLDs (site.es) — pros/cons of each
+     - Canonical tag handling for similar content across languages
+     - XML sitemap structure for multilingual content
+
+  4. Content production workflow:
+     - Native speakers vs AI translation vs professional translation
+     - SEO review process per language
+     - How to maintain consistency across languages
+     - Publishing cadence for 4 language versions
+
+  5. Common international SEO mistakes:
+     - Auto-translation without keyword research
+     - Same content for Spain Spanish and Latin American Spanish
+     - Missing hreflang tags causing wrong language pages to rank
+     - Ignoring local search engines (not just Google)
+
+assertions:
+  - type: llm_judge
+    criteria: "The keyword localization demonstrates genuine understanding — shows that direct translation misses actual search behavior, with specific examples of how people in different countries search differently for the same concept"
+    weight: 0.30
+    description: "Genuine keyword localization understanding"
+  - type: llm_judge
+    criteria: "The hreflang implementation is technically correct — proper tag syntax, correct language-region codes (es-ES vs es-MX), and handles the x-default tag for fallback"
+    weight: 0.25
+    description: "Technically correct hreflang implementation"
+  - type: llm_judge
+    criteria: "The URL structure recommendation is well-reasoned — explains trade-offs between subdirectories (easier management) and ccTLDs (stronger local signal) with a clear recommendation based on the scenario"
+    weight: 0.25
+    description: "Well-reasoned URL structure recommendation"
+  - type: llm_judge
+    criteria: "The Spain vs Latin America distinction is addressed — correctly notes that Spanish content needs separate versions for different regional markets with different keywords, idioms, and financial contexts"
+    weight: 0.20
+    description: "Correct regional Spanish distinction"

package/courses/blog-post-seo-writing/scenarios/level-3/programmatic-seo-content.yaml

@@ -0,0 +1,76 @@
+meta:
+  id: programmatic-seo-content
+  level: 3
+  course: blog-post-seo-writing
+  type: output
+  description: "Design programmatic SEO strategies for scalable content creation"
+  tags: [seo, programmatic-seo, scalable-content, templates, automation]
+
+state: {}
+
+trigger: |
+  Programmatic SEO creates hundreds or thousands of pages from templates
+  and data. Companies like Zapier (integration pages), NomadList (city pages),
+  and Wise (currency converter pages) drive millions of visits this way.
+
+  You're building programmatic SEO for a job board platform.
+
+  Design the programmatic content system:
+
+  1. Page type identification:
+     Design 3 programmatic page types:
+     a) "[Job Title] Salary in [City]" pages (e.g., "Software Engineer
+        Salary in Austin")
+        - Data sources needed
+        - Template structure
+        - Unique content elements per page
+        - Estimated page count: ~5,000 pages
+
+     b) "[Job Title] Job Description Template" pages
+        - Template structure with variable sections
+        - How each page is unique (not just city swapped)
+        - Estimated page count: ~500 pages
+
+     c) "[Company] Reviews and Salaries" pages
+        - Data integration requirements
+        - User-generated content component
+        - Estimated page count: ~10,000 pages
+
+  2. Template design (for the salary page type):
+     - Write the full page template with variables marked
+     - H1, meta title, meta description templates
+     - Body content sections (mix of templated + dynamic data)
+     - How to make each page genuinely useful (not thin/doorway)
+     - Minimum unique content threshold per page
+
+  3. Quality at scale:
+     - How to avoid Google's "thin content" penalty with templated pages
+     - Minimum quality bar per page
+     - When to noindex low-value pages (low search volume locations)
+     - Internal linking strategy across thousands of pages
+     - Sitemap management for large page sets
+
+  4. Risks and mitigation:
+     - Doorway page penalty risk
+     - Index bloat and crawl budget
+     - Duplicate content across similar pages
+     - How Google's helpful content system evaluates programmatic content
+     - When programmatic SEO is appropriate vs when it's spam
+
+assertions:
+  - type: llm_judge
+    criteria: "The page templates create genuinely unique pages — not just swapping a city name but including unique data, insights, and context that make each page useful on its own"
+    weight: 0.30
+    description: "Genuinely unique templated pages"
+  - type: llm_judge
+    criteria: "The quality-at-scale strategy is realistic — addresses the real tension between creating thousands of pages and maintaining quality, with specific thresholds and noindex criteria"
+    weight: 0.25
+    description: "Realistic quality-at-scale strategy"
+  - type: llm_judge
+    criteria: "The risk assessment correctly identifies doorway page penalties — explains what makes programmatic SEO legitimate (unique value per page) vs spammy (thin template pages with swapped variables)"
+    weight: 0.25
+    description: "Correct doorway page risk assessment"
+  - type: llm_judge
+    criteria: "The internal linking strategy handles scale — explains how to link across thousands of pages without creating an unmanageable link structure, using hub pages, breadcrumbs, and related page suggestions"
+    weight: 0.20
+    description: "Scalable internal linking strategy"

package/courses/blog-post-seo-writing/scenarios/level-3/seo-ab-testing.yaml

@@ -0,0 +1,74 @@
+meta:
+  id: seo-ab-testing
+  level: 3
+  course: blog-post-seo-writing
+  type: output
+  description: "Design and interpret A/B tests for SEO content elements"
+  tags: [seo, ab-testing, experimentation, data-driven, optimization]
+
+state: {}
+
+trigger: |
+  SEO A/B testing lets you make data-driven decisions about content changes
+  instead of relying on best practices alone. But SEO testing is different
+  from conversion testing — you can't split traffic at the Google level.
+
+  You're the SEO manager for a large e-commerce blog (500+ posts).
+
+  Design SEO A/B tests for these elements:
+
+  1. Title tag test:
+     - Hypothesis: Adding the year to title tags increases CTR
+     - Test group: 50 product review posts
+     - Control: "Best Running Shoes (Expert Reviewed)"
+     - Variant: "Best Running Shoes 2026 (Expert Reviewed)"
+     - How to measure: CTR from Search Console
+     - Duration: minimum test period and why
+     - How to determine statistical significance for SEO tests
+
+  2. Content length test:
+     - Hypothesis: Expanding thin content from 800 to 2,000 words
+       improves rankings
+     - Test group: 30 posts currently at 600-900 words
+     - What to expand (not just fluff — substantive additions)
+     - How to isolate the length variable from other changes
+     - Measurement: position change, traffic change, engagement metrics
+
+  3. Internal linking test:
+     - Hypothesis: Adding 5 contextual internal links per post improves
+       linked page rankings
+     - Test group: 40 posts with 0-1 internal links
+     - Control: no changes
+     - Variant: add 5 relevant internal links
+     - How to measure the impact on BOTH the linking and linked pages
+
+  4. Content structure test:
+     - Hypothesis: Adding a TL;DR summary at the top improves engagement
+       and reduces bounce rate
+     - How to implement without hurting featured snippet targeting
+     - Metrics to track
+
+  For each test:
+  - Clear hypothesis and expected outcome
+  - Test design (groups, duration, sample size)
+  - Measurement methodology
+  - How to account for SEO confounding variables
+  - Decision criteria: when to roll out vs roll back
+
+assertions:
+  - type: llm_judge
+    criteria: "The testing methodology accounts for SEO-specific challenges — Google doesn't allow traditional traffic splitting, so the approach uses time-based or page-group-based testing with appropriate controls"
+    weight: 0.30
+    description: "SEO-specific testing methodology"
+  - type: llm_judge
+    criteria: "Statistical significance is addressed correctly — SEO tests need longer durations (4-8 weeks minimum) due to Google's indexing and ranking update cycles, and this is explained"
+    weight: 0.25
+    description: "Correct statistical significance approach"
+  - type: llm_judge
+    criteria: "Confounding variables are identified — seasonality, algorithm updates, competitor changes, and other ranking factors that could skew results are acknowledged with mitigation strategies"
+    weight: 0.25
+    description: "Confounding variables identified"
+  - type: llm_judge
+    criteria: "Decision criteria are clear — specific thresholds for when to declare a test successful, when to extend, and when to roll back changes"
+    weight: 0.20
+    description: "Clear decision criteria"

package/courses/blog-post-seo-writing/scenarios/level-3/seo-content-audit.yaml

@@ -0,0 +1,76 @@
+meta:
+  id: seo-content-audit
+  level: 3
+  course: blog-post-seo-writing
+  type: output
+  description: "Audit a blog's content library to identify what to update, merge, or delete"
+  tags: [seo, content-audit, content-pruning, optimization, strategy]
+
+state: {}
+
+trigger: |
+  A content audit identifies which posts to update, merge, redirect, or
+  delete. Most blogs have 30-40% of content that's hurting overall site
+  quality and cannibalizing rankings.
+
+  You're auditing a SaaS company blog with 200 published posts over 4 years.
+
+  Here's a sample of 10 posts with their metrics:
+
+  Post 1: "What Is CRM?" — 15,000 visits/mo, position #4, 85 backlinks
+  Post 2: "CRM Software Guide" — 3,200 visits/mo, position #8, 42 backlinks
+  Post 3: "What Is CRM Software?" — 400 visits/mo, position #28, 5 backlinks
+  Post 4: "Best CRM for Small Business 2023" — 800 visits/mo, position #14, 22 backlinks
+  Post 5: "CRM Implementation Steps" — 1,100 visits/mo, position #6, 18 backlinks
+  Post 6: "CRM vs Spreadsheets" — 200 visits/mo, position #35, 2 backlinks
+  Post 7: "Our Company Update — New Office" — 15 visits/mo, no ranking, 0 backlinks
+  Post 8: "Top 10 CRM Features" — 900 visits/mo, position #11, 12 backlinks
+  Post 9: "CRM for Real Estate" — 2,400 visits/mo, position #3, 30 backlinks
+  Post 10: "CRM Pricing Comparison" — 1,800 visits/mo, position #7, 25 backlinks
+
+  Perform the full audit:
+
+  1. Categorize each post (with reasoning):
+     - Keep as-is (performing well)
+     - Update (has potential but needs refresh)
+     - Merge (overlapping with another post)
+     - Redirect & delete (low value, hurting site quality)
+     - Repurpose (wrong format but good topic)
+
+  2. Identify cannibalization issues:
+     - Which posts are competing for the same keywords?
+     - How to resolve: merge, differentiate, or redirect?
+     - The correct redirect strategy (301 to the winner)
+
+  3. Content consolidation plan:
+     - Which posts should be merged and into what?
+     - How to preserve backlinks during merges (redirect strategy)
+     - Expected traffic impact of consolidation
+
+  4. Update priority matrix:
+     - Rank posts by update priority
+     - Scoring: (current traffic + backlinks + keyword potential) / update effort
+     - Which posts give the highest ROI for update effort?
+
+  5. Blog health score:
+     - What percentage of these posts are healthy vs unhealthy?
+     - Extrapolate to the full 200-post blog
+     - Recommended audit frequency
+
+assertions:
+  - type: llm_judge
+    criteria: "Posts 1, 2, and 3 are correctly identified as a cannibalization cluster — three posts targeting variations of 'what is CRM / CRM software' should be consolidated, with the strongest post (Post 1) as the winner"
+    weight: 0.30
+    description: "Correctly identified cannibalization cluster"
+  - type: llm_judge
+    criteria: "Post 7 (company update) is correctly flagged for deletion or noindex — it has no SEO value, no backlinks, and dilutes crawl budget and site quality"
+    weight: 0.20
+    description: "Correctly flagged low-value content"
+  - type: llm_judge
+    criteria: "The redirect strategy preserves link equity — when merging posts, 301 redirects from deleted posts to the consolidated post are specified, and backlink preservation is addressed"
+    weight: 0.25
+    description: "Link equity preserving redirect strategy"
+  - type: llm_judge
+    criteria: "The priority matrix correctly identifies Post 4 as a high-ROI update target — it has decent backlinks and traffic but is outdated (2023), making a year update likely to recover significant rankings"
+    weight: 0.25
+    description: "Correct high-ROI update identification"

package/courses/blog-post-seo-writing/scenarios/level-4/advanced-serp-analysis.yaml

@@ -0,0 +1,83 @@
+meta:
+  id: advanced-serp-analysis
+  level: 4
+  course: blog-post-seo-writing
+  type: output
+  description: "Analyze SERPs deeply to reverse-engineer Google's ranking preferences"
+  tags: [seo, serp-analysis, ranking-factors, competitive-analysis, advanced]
+
+state: {}
+
+trigger: |
+  Advanced SERP analysis goes beyond "check what's ranking" — it reverse-
+  engineers WHY certain content ranks to identify the exact content
+  attributes Google prefers for specific queries.
+
+  Perform deep SERP analysis for 3 different query types:
+
+  1. Informational query: "how does machine learning work"
+     Analyze the hypothetical top 5 results:
+     - #1: University .edu comprehensive guide (8,000 words)
+     - #2: Tech publication beginner's guide (3,000 words)
+     - #3: YouTube video (embedded in search)
+     - #4: Company blog with interactive examples (2,500 words)
+     - #5: Wikipedia article
+
+     Analyze:
+     - Why does the .edu outrank the tech publication?
+     - What content attributes are common across all top 5?
+     - What SERP features appear (AI Overview, video carousel,
+       People Also Ask, knowledge panel)?
+     - What content type should you create to rank here?
+     - Content blueprint: specific structure, depth, and format needed
+
+  2. Commercial query: "best CRM software for small business"
+     Analyze hypothetical top 5:
+     - #1: G2 comparison page with user reviews
+     - #2: Forbes Advisor listicle (updated monthly)
+     - #3: PCMag comprehensive review (expert-tested)
+     - #4: Capterra comparison with pricing
+     - #5: Individual blog with personal experience
+
+     Analyze:
+     - Why do review aggregators dominate this query?
+     - Can a single-author blog compete? What angle?
+     - Commercial intent signals Google looks for
+     - How to differentiate when 4/5 results are similar listicles
+
+  3. Navigational/transactional query: "Salesforce pricing 2026"
+     Analyze hypothetical top 5:
+     - #1: Salesforce.com official pricing page
+     - #2: Salesforce pricing comparison blog post
+     - #3: G2 Salesforce pricing and plans
+     - #4: Blog: "Salesforce Pricing Explained (Hidden Costs)"
+     - #5: YouTube: Salesforce pricing breakdown
+
+     Analyze:
+     - Why does the official page rank first?
+     - What angle lets blogs compete for branded queries?
+     - The "hidden costs" and "explained" angles — why they work
+     - How to create content that adds value beyond the official page
+
+  For each analysis provide:
+  - Content attribute comparison table (word count, format, freshness, authority, E-E-A-T signals)
+  - The "ranking formula" Google seems to use for this query type
+  - Specific content blueprint to compete
+
+assertions:
+  - type: llm_judge
+    criteria: "The analysis identifies different ranking factors per query type — informational queries prioritize depth and authority, commercial queries prioritize comprehensiveness and freshness, navigational queries prioritize brand authority and unique angles"
+    weight: 0.30
+    description: "Distinct ranking factors per query type"
+  - type: llm_judge
+    criteria: "The competitive analysis is nuanced — explains why specific result types outrank others (e.g., .edu domain authority for informational, review aggregators' data for commercial) rather than generic authority claims"
+    weight: 0.25
+    description: "Nuanced competitive ranking analysis"
+  - type: llm_judge
+    criteria: "The content blueprints are actionable — not just 'write comprehensive content' but specific structure, word count ranges, content elements, and differentiation strategies for each query type"
+    weight: 0.25
+    description: "Actionable content blueprints"
+  - type: llm_judge
+    criteria: "SERP features are analyzed — identifies which features appear for each query type (AI Overviews, video carousels, PAA, knowledge panels) and how they affect the organic click-through opportunity"
+    weight: 0.20
+    description: "SERP feature impact analysis"

package/courses/blog-post-seo-writing/scenarios/level-4/ai-overview-optimization.yaml

@@ -0,0 +1,79 @@
+meta:
+  id: ai-overview-optimization
+  level: 4
+  course: blog-post-seo-writing
+  type: output
+  description: "Optimize blog content for Google's AI Overviews and generative search"
+  tags: [seo, ai-overviews, generative-search, sge, future-seo]
+
+state: {}
+
+trigger: |
+  Google's AI Overviews (formerly SGE) generate AI-powered summaries at
+  the top of search results, citing sources. This changes SEO fundamentally:
+  your content needs to be good enough to be cited by AI, not just ranked.
+
+  In 2026, AI Overviews appear for roughly 40% of informational queries
+  and are expanding to more query types.
+
+  You're adapting a technology media site's SEO strategy for AI Overviews.
+
+  Provide the complete AI Overview optimization strategy:
+
+  1. How AI Overviews change SEO:
+     - Which query types trigger AI Overviews (and which don't)
+     - How AI Overviews select and cite sources
+     - Impact on organic CTR (data and estimates for 2026)
+     - Which content types benefit from AI Overviews (getting cited)
+     - Which content types are hurt (traffic stolen by the summary)
+
+  2. Content optimization for AI citation:
+     - What makes content "citable" by Google's AI
+     - Structural elements that increase citation probability
+     - Writing style that AI summaries prefer to quote
+     - How to format data, statistics, and expert opinions for citation
+     - Write a 300-word blog section optimized for AI citation on the
+       topic "What is quantum computing?" — annotate why each element
+       is optimized for AI citation
+
+  3. Query classification framework:
+     Classify these queries by AI Overview impact:
+     a) "what is kubernetes" — full AI Overview likely
+     b) "best CRM software 2026" — partial AI Overview with product links
+     c) "how to configure nginx reverse proxy" — code-focused, less likely
+     d) "is Salesforce worth the price" — opinion-based, variable
+     e) "[specific error code] fix" — AI Overview with solution steps
+
+     For each: expected AI Overview behavior, impact on organic traffic,
+     and content strategy adaptation
+
+  4. Defensive strategies:
+     - How to maintain traffic when AI Overviews answer the query
+     - Creating content that goes beyond what AI can summarize
+     - Building brand recognition so users click through anyway
+     - Long-tail and niche queries where AI Overviews don't appear
+     - Original research and unique data that AI must cite
+
+  5. Measurement and adaptation:
+     - How to track AI Overview citations
+     - Metrics that matter in an AI Overview world
+     - When to pivot content strategy based on AI Overview changes
+     - Future-proofing: what happens as AI Overviews expand further
+
+assertions:
+  - type: llm_judge
+    criteria: "The AI Overview analysis is current and accurate — correctly represents 2026 AI Overview behavior, including which queries trigger them, how sources are selected, and the real impact on CTR"
+    weight: 0.30
+    description: "Current accurate AI Overview analysis"
+  - type: llm_judge
+    criteria: "The citation-optimized writing sample is clearly annotated — each element (clear definitions, structured data, expert attribution, specific statistics) is explained in terms of why it increases AI citation probability"
+    weight: 0.25
+    description: "Clearly annotated citation-optimized writing"
+  - type: llm_judge
+    criteria: "The defensive strategies are substantive — goes beyond 'write better content' to specific tactics like creating content AI can't summarize, building brand recognition, and targeting AI Overview-free query types"
+    weight: 0.25
+    description: "Substantive defensive strategies"
+  - type: llm_judge
+    criteria: "The query classification is nuanced — different query types get different AI Overview treatments, and the content strategy adapts accordingly rather than applying one approach to all queries"
+    weight: 0.20
+    description: "Nuanced query classification"

package/courses/blog-post-seo-writing/scenarios/level-4/eeat-author-building.yaml

@@ -0,0 +1,76 @@
+meta:
+  id: eeat-author-building
+  level: 4
+  course: blog-post-seo-writing
+  type: output
+  description: "Build comprehensive E-E-A-T author profiles that boost content rankings"
+  tags: [seo, eeat, author-authority, expertise, trust-signals]
+
+state: {}
+
+trigger: |
+  Google's E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness)
+  framework heavily weights WHO writes the content. Author entities with
+  verifiable expertise rank better, especially for YMYL topics.
+
+  You're building E-E-A-T author profiles for a health and wellness
+  publication with 5 authors:
+
+  Author 1: Dr. Lisa Park — Board-certified dermatologist, 15 years practice
+  Author 2: Mike Torres — Certified personal trainer, NASM-CPT, 8 years
+  Author 3: Sarah Kim — Registered dietitian, MS Nutrition, 6 years
+  Author 4: Alex Chen — Health journalist, no medical credentials, 10 years writing
+  Author 5: Jordan Wright — New hire, recently graduated nutrition science degree
+
+  For each author, design their E-E-A-T optimization:
+
+  1. Author page design:
+     - What to include on the author bio page
+     - Credentials display (degrees, certifications, affiliations)
+     - Experience signals (years of practice, patient count, publications)
+     - External verification links (medical board, LinkedIn, publications)
+     - Author schema markup (Person schema with credentials)
+
+  2. Cross-platform authority signals:
+     - Which external platforms each author should have profiles on
+     - How Google connects author profiles across the web
+     - Publication history and citation building
+     - Speaking engagements, podcast appearances, media quotes
+
+  3. Content-level E-E-A-T signals:
+     - How authors should demonstrate Experience in their writing
+       (first-person clinical examples for Dr. Park, client stories for Mike)
+     - Expert review badges for content reviewed by credentialed authors
+     - Medical review disclosure format for health content
+     - How to handle Author 4 (journalist) writing YMYL content
+       (expert review process, disclosure, co-authoring strategies)
+
+  4. Building E-E-A-T for Author 5 (new hire):
+     - How to build authority for someone without established credentials
+     - Co-authoring strategy with senior authors
+     - Content types appropriate for their experience level
+     - 6-month E-E-A-T building roadmap
+
+  5. E-E-A-T audit checklist:
+     - How to audit existing content for E-E-A-T signals
+     - Common E-E-A-T deficiencies and fixes
+     - Priority: which content needs E-E-A-T upgrades most urgently
+     - How to measure E-E-A-T improvement impact on rankings
+
+assertions:
+  - type: llm_judge
+    criteria: "The author page designs include verifiable credentials — not just self-reported claims but links to medical boards, certification databases, and external publications that Google can verify"
+    weight: 0.25
+    description: "Verifiable credential display"
+  - type: llm_judge
+    criteria: "The journalist (Author 4) YMYL strategy is sound — correctly addresses that a non-credentialed writer needs expert review processes, medical disclaimers, and co-authoring with credentialed experts for health content"
+    weight: 0.25
+    description: "Sound journalist YMYL strategy"
+  - type: llm_judge
+    criteria: "The new author (Author 5) roadmap is realistic — building E-E-A-T takes time, and the plan acknowledges this with appropriate content restrictions early on and gradual authority building"
+    weight: 0.25
+    description: "Realistic new author E-E-A-T roadmap"
+  - type: llm_judge
+    criteria: "The content-level Experience signals are specific — shows how each author type can demonstrate first-hand experience (clinical cases, client transformations, dietary consultations) rather than generic expertise claims"
+    weight: 0.25
+    description: "Specific Experience signal demonstrations"

package/courses/blog-post-seo-writing/scenarios/level-4/entity-based-seo.yaml

@@ -0,0 +1,76 @@
+meta:
+  id: entity-based-seo
+  level: 4
+  course: blog-post-seo-writing
+  type: output
+  description: "Optimize blog content for Google's entity-based understanding and Knowledge Graph"
+  tags: [seo, entities, knowledge-graph, semantic-seo, advanced]
+
+state: {}
+
+trigger: |
+  Google's Knowledge Graph understands entities (people, places, concepts,
+  brands) and their relationships. Entity-based SEO means writing content
+  that helps Google connect your content to its entity understanding,
+  improving relevance signals beyond keywords.
+
+  You're optimizing a healthcare technology blog for entity-based SEO.
+
+  Provide the complete entity optimization strategy:
+
+  1. Entity identification for your niche:
+     - Map the key entities in healthcare technology
+     - Primary entities: your brand, products, key personnel
+     - Related entities: technologies, conditions, regulations, organizations
+     - Show the entity relationship map (brand → products → use cases →
+       conditions → regulations)
+     - How Google's Knowledge Graph connects these entities
+
+  2. Content optimization for entities:
+     Take this blog post topic: "How Telehealth Is Transforming
+     Chronic Disease Management"
+
+     - Identify all entities in the topic (telehealth, chronic disease,
+       specific conditions, technologies, regulatory bodies)
+     - Write a 500-word section that naturally incorporates entity
+       relationships
+     - Show how to use co-occurring entities to strengthen topical relevance
+     - Demonstrate entity-first writing vs keyword-first writing
+
+  3. Author entity building:
+     - How to establish your authors as recognized entities
+     - Google Scholar, Wikipedia, professional profiles
+     - Structured data for author entities (Person schema)
+     - Why author entities matter more post-E-E-A-T update
+     - Building author pages that Google can connect to external profiles
+
+  4. Brand entity optimization:
+     - How to get your brand into Google's Knowledge Graph
+     - Knowledge panel triggering strategies
+     - Consistent entity signals across the web (Wikipedia, Wikidata,
+       Crunchbase, LinkedIn)
+     - How brand entity strength affects all content rankings
+
+  5. Entity-based content planning:
+     - How to use entity research to find content topics
+     - Tools and methods for entity analysis
+     - Building content that fills entity relationship gaps
+     - Measuring entity-based SEO progress
+
+assertions:
+  - type: llm_judge
+    criteria: "The entity relationship map is coherent — shows how healthcare technology entities connect to each other in ways that reflect Google's actual Knowledge Graph structure"
+    weight: 0.25
+    description: "Coherent entity relationship map"
+  - type: llm_judge
+    criteria: "The entity-optimized writing sample demonstrates the difference — shows how mentioning related entities (specific conditions, FDA, specific telehealth platforms) strengthens topical relevance compared to keyword-only optimization"
+    weight: 0.30
+    description: "Clear entity vs keyword writing difference"
+  - type: llm_judge
+    criteria: "The author entity strategy is actionable — specific steps for building author recognition that Google can verify (not just 'create a bio page' but linking to external authority sources)"
+    weight: 0.25
+    description: "Actionable author entity strategy"
+  - type: llm_judge
+    criteria: "The Knowledge Graph entry strategy is realistic — accurately represents how brands enter the Knowledge Graph (consistent structured data, Wikipedia notability, Wikidata entries) rather than oversimplifying"
+    weight: 0.20
+    description: "Realistic Knowledge Graph entry strategy"