{"slug":"koray-content-audit","title":"Koray Content Audit — Semantic Alignment Verification","tags":["seo","semantic","koray","content-audit","discourse-integration"],"agent_summary":"Audits existing articles for semantic alignment with their brief: relevance, discourse integration, distributional semantics, contextual coverage balance, and entity completeness.","trigger_phrases":["content audit Koray","semantic audit existing content","discourse integration audit","distributional semantics check","article quality audit Koray","content alignment audit","contextual coverage audit"],"runnable":true,"markdown":"\n## Core Concept: What Is a Koray Content Audit?\n\nA content audit verifies that every written article ALIGNS with its content brief, contextual vector, and query network. This is NOT a grammar check, readability score, or Flesch-Kincaid test. It is semantic alignment verification: does the article do what the brief instructed, in the correct order, with the correct entities, using the correct answer formats, with proper discourse integration?\n\nAudit a page when:\n- Rankings dropped after a core algorithm update\n- A new article was written and needs review before publishing\n- A competitor significantly outranks on a similar topic\n- The content team delivered work against a brief\n\n## The Seven Audit Dimensions\n\n### 1. Brief Responsiveness\nDoes the article address every section in the brief? Check heading by heading. Score each section: Covered / Partially covered / Missing.\n\nMissing sections are the highest-priority fix. No amount of writing quality compensates for a section gap.\n\n### 2. Contextual Vector Alignment\nDoes the heading structure match the brief's specified contextual vector? Check:\n- H1 contains Central Entity + attribute\n- H2s appear in the specified order\n- H3s are question-format where specified in brief\n- No extra headings added that shift topical focus\n\nAny deviation must be justified by evidence that the changed order improves intent fulfillment.\n\n### 3. Discourse Integration Score\nRead each paragraph as a narrative. For each sentence transition, ask: does sentence N+1 logically follow from sentence N?\n\nScore: Strong (clear logical flow throughout) / Adequate (occasional gaps) / Broken (sentences that could be shuffled without loss of meaning).\n\nBroken discourse is the most common quality issue in AI-generated and outsourced content. Rewrite from the first broken transition.\n\n### 4. Distributional Semantics Check\nExtract all context terms from the brief. For each context term, find where it appears in the article. Map against expected sections. Flag:\n- Terms only appearing in one location (should be distributed)\n- Entire sections with no context terms (likely generic filler)\n- Front-loaded articles where all context terms appear in the first 30%\n\n### 5. Answer Format Compliance\nFor each section, verify the format matches the brief specification:\n- Definition format: starts with entity, 25-50 words, no qualifiers\n- Procedure format: numbered list, action-verb first per step\n- Comparison format: table or parallel structure with explicit differentiator\n- Value format: numeric range in first sentence with units\n\n### 6. Entity Completeness\nCount entity occurrences vs. brief's minimum entity checklist. Flag:\n- Required entities that appear 0 times\n- Required entities that appear only in one section (should be distributed)\n- Forbidden entities that were included\n\n### 7. Featured Snippet Delivery\nFor each H3 marked as a featured snippet target in the brief:\n- Does a direct answer appear in the first 60 words after the heading?\n- Is the answer declarative (no qualifiers like \"typically\", \"usually\", \"may\")?\n- Is the answer standalone-extractable (makes sense without surrounding context)?\n\n## Audit Output Format\n\nDeliver audit as a scored table:\n\n| Dimension | Score (1-5) | Issues Found | Priority Fixes |\n|-----------|------------|-------------|---------------|\n| Brief Responsiveness | X | X missing sections | List sections |\n| Contextual Vector | X | X deviations | List deviations |\n| Discourse Integration | X | X broken transitions | Paragraph locations |\n| Distributional Semantics | X | X front-loaded terms | Term list |\n| Answer Formats | X | X non-compliant sections | Section list |\n| Entity Completeness | X | X missing entities | Entity list |\n| Featured Snippet | X | X failed targets | H3 list |\n\nTotal audit score: X/35. Below 25 = full rewrite. 25-30 = targeted revision. 30-35 = minor polish.\n\n## Cross-References\n\n- `koray-content-brief` — the brief that the audit checks against\n- `koray-microsemantics` — reference for answer format definitions during scoring\n- `koray-contextual-vector` — heading structure standard for dimension 2\n\n#seo-sop #seo #semantic #koray #content-audit #discourse-integration\n","html":"<h2>Core Concept: What Is a Koray Content Audit?</h2>\n<p>A content audit verifies that every written article ALIGNS with its content brief, contextual vector, and query network. This is NOT a grammar check, readability score, or Flesch-Kincaid test. It is semantic alignment verification: does the article do what the brief instructed, in the correct order, with the correct entities, using the correct answer formats, with proper discourse integration?</p>\n<p>Audit a page when:</p>\n<ul>\n<li>Rankings dropped after a core algorithm update</li>\n<li>A new article was written and needs review before publishing</li>\n<li>A competitor significantly outranks on a similar topic</li>\n<li>The content team delivered work against a brief</li>\n</ul>\n<h2>The Seven Audit Dimensions</h2>\n<h3>1. Brief Responsiveness</h3>\n<p>Does the article address every section in the brief? Check heading by heading. Score each section: Covered / Partially covered / Missing.</p>\n<p>Missing sections are the highest-priority fix. No amount of writing quality compensates for a section gap.</p>\n<h3>2. Contextual Vector Alignment</h3>\n<p>Does the heading structure match the brief's specified contextual vector? Check:</p>\n<ul>\n<li>H1 contains Central Entity + attribute</li>\n<li>H2s appear in the specified order</li>\n<li>H3s are question-format where specified in brief</li>\n<li>No extra headings added that shift topical focus</li>\n</ul>\n<p>Any deviation must be justified by evidence that the changed order improves intent fulfillment.</p>\n<h3>3. Discourse Integration Score</h3>\n<p>Read each paragraph as a narrative. For each sentence transition, ask: does sentence N+1 logically follow from sentence N?</p>\n<p>Score: Strong (clear logical flow throughout) / Adequate (occasional gaps) / Broken (sentences that could be shuffled without loss of meaning).</p>\n<p>Broken discourse is the most common quality issue in AI-generated and outsourced content. Rewrite from the first broken transition.</p>\n<h3>4. Distributional Semantics Check</h3>\n<p>Extract all context terms from the brief. For each context term, find where it appears in the article. Map against expected sections. Flag:</p>\n<ul>\n<li>Terms only appearing in one location (should be distributed)</li>\n<li>Entire sections with no context terms (likely generic filler)</li>\n<li>Front-loaded articles where all context terms appear in the first 30%</li>\n</ul>\n<h3>5. Answer Format Compliance</h3>\n<p>For each section, verify the format matches the brief specification:</p>\n<ul>\n<li>Definition format: starts with entity, 25-50 words, no qualifiers</li>\n<li>Procedure format: numbered list, action-verb first per step</li>\n<li>Comparison format: table or parallel structure with explicit differentiator</li>\n<li>Value format: numeric range in first sentence with units</li>\n</ul>\n<h3>6. Entity Completeness</h3>\n<p>Count entity occurrences vs. brief's minimum entity checklist. Flag:</p>\n<ul>\n<li>Required entities that appear 0 times</li>\n<li>Required entities that appear only in one section (should be distributed)</li>\n<li>Forbidden entities that were included</li>\n</ul>\n<h3>7. Featured Snippet Delivery</h3>\n<p>For each H3 marked as a featured snippet target in the brief:</p>\n<ul>\n<li>Does a direct answer appear in the first 60 words after the heading?</li>\n<li>Is the answer declarative (no qualifiers like \"typically\", \"usually\", \"may\")?</li>\n<li>Is the answer standalone-extractable (makes sense without surrounding context)?</li>\n</ul>\n<h2>Audit Output Format</h2>\n<p>Deliver audit as a scored table:</p>\n<p>| Dimension | Score (1-5) | Issues Found | Priority Fixes |\n|-----------|------------|-------------|---------------|\n| Brief Responsiveness | X | X missing sections | List sections |\n| Contextual Vector | X | X deviations | List deviations |\n| Discourse Integration | X | X broken transitions | Paragraph locations |\n| Distributional Semantics | X | X front-loaded terms | Term list |\n| Answer Formats | X | X non-compliant sections | Section list |\n| Entity Completeness | X | X missing entities | Entity list |\n| Featured Snippet | X | X failed targets | H3 list |</p>\n<p>Total audit score: X/35. Below 25 = full rewrite. 25-30 = targeted revision. 30-35 = minor polish.</p>\n<h2>Cross-References</h2>\n<ul>\n<li><code>koray-content-brief</code> — the brief that the audit checks against</li>\n<li><code>koray-microsemantics</code> — reference for answer format definitions during scoring</li>\n<li><code>koray-contextual-vector</code> — heading structure standard for dimension 2</li>\n</ul>\n<p>#seo-sop #seo #semantic #koray #content-audit #discourse-integration</p>\n"}