dpaluy · June 11, 2025 04:09 · dpaluy · Jun 11, 2025
diff --git a/content-production.mdc b/content-production.mdc
 1 Narrative Structure
 1.  Three-Act Layout, Not H2 Soup
 • Act I (≈10 %)—Hook + keyword-rich promise.
 • Act II (≈75 %)—Deep exploration broken into 3-5 logical “beats” (single bold lead-in line okay; avoid nested H3s/H4s).
 • Act III (≈15 %)—Synthesize insights, call-to-action, and next-step links.
 2.  Natural Transitions
 • Bridge sections with “So what?” sentences instead of new headings.
 • Use connective devices (questions, contrasts, mini-stories) every 300-400 words to keep readers moving.
 3.  Strategic Subheadings Only
 • H2 count cap: max 5 for the entire article.
 • No H3/H4 unless absolutely required for step-by-step clarity.

 2 Keyword & Entity Handling
 1.  Primary Keyword Placement
 • Title, first 90 characters, one H2, meta description, and final paragraph.
 2.  Density Discipline
 • Target 0.8 %-1 % primary; 0.3 %-0.6 % for each secondary.
 3.  Contextual Entities
 • Integrate 6-10 semantically related entities (people, brands, concepts) that appear in top-ranking SERP pages—spread organically.
 4.  Zero-Force Rule
 • Rewrite awkward exact matches; grammar beats rigidity.

 3 Engagement & Readability
 1.  Mixed Paragraph Rhythm—2-4 sentences average; a few one-liners permitted for punch.
 2.  Story Inserts—Include one concrete anecdote, case study, or stat box every ~600 words.
 3.  Embedded Media—Recommend 1 infographic, chart, or image per 800-1,000 words with alt text (“describe + keyword”).
 4.  Sensory + Action Verbs—Aim for ≥25 % sentences containing vivid adjectives or verbs to humanize tone.
 5.  It's okay to occasionally be a little light-hearted/approachable. Be authoritative, but friendly (however do NOT be cheesy).

 4 LLM-Specific Instructions
 1.  Token Variety Monitor—No identical sentence stems within 5 sentences; vary n-grams.
 2.  Hallucination Guardrails—Cite at least one reputable source per Act II beat; flag uncertain facts with “VERIFY:”.
 3.  Self-Revision Prompt—After draft, run revise_for_repetition_and_clarity() to trim fluff and merge redundant fragments.
 4.  Temperature Guideline—0.7 for creative hooks; 0.3 for definition sections.

 5 Linking
 1.  Internal Links—3-6 pointing to related cornerstone pages; avoid exact-match anchors on every link.

 6 Meta & Schema
 1.  FAQ-Schema Block—Optional; limit to 3 high-intent Q&As appended after Act III.

 7 Pre-Publish Checklist
 • ❑ Opening 40 words answer search intent.
 • ❑ H2 count ≤ 5; no orphan H3/H4.
 • ❑ Keyword densities in range; reads naturally aloud.
 • ❑ Each Act II beat contains one credible stat or expert quote.
 • ❑ Flesch Reading Ease ≥ 60.
 • ❑ search the @⁠Web for 3-5 sources that you can reference and link to in the article, then write the article based on keyword research.
diff --git a/Prompt.md b/Prompt.md
	1 Narrative Structure
	1. Three-Act Layout, Not H2 Soup
	• Act I (≈10 %)—Hook + keyword-rich promise.
	• Act II (≈75 %)—Deep exploration broken into 3-5 logical “beats” (single bold lead-in line okay; avoid nested H3s/H4s).
	• Act III (≈15 %)—Synthesize insights, call-to-action, and next-step links.
	2. Natural Transitions
	• Bridge sections with “So what?” sentences instead of new headings.
	• Use connective devices (questions, contrasts, mini-stories) every 300-400 words to keep readers moving.
	3. Strategic Subheadings Only
	• H2 count cap: max 5 for the entire article.
	• No H3/H4 unless absolutely required for step-by-step clarity.

	2 Keyword & Entity Handling
	1. Primary Keyword Placement
	• Title, first 90 characters, one H2, meta description, and final paragraph.
	2. Density Discipline
	• Target 0.8 %-1 % primary; 0.3 %-0.6 % for each secondary.
	3. Contextual Entities
	• Integrate 6-10 semantically related entities (people, brands, concepts) that appear in top-ranking SERP pages—spread organically.
	4. Zero-Force Rule
	• Rewrite awkward exact matches; grammar beats rigidity.

	3 Engagement & Readability
	1. Mixed Paragraph Rhythm—2-4 sentences average; a few one-liners permitted for punch.
	2. Story Inserts—Include one concrete anecdote, case study, or stat box every ~600 words.
	3. Embedded Media—Recommend 1 infographic, chart, or image per 800-1,000 words with alt text (“describe + keyword”).
	4. Sensory + Action Verbs—Aim for ≥25 % sentences containing vivid adjectives or verbs to humanize tone.
	5. It's okay to occasionally be a little light-hearted/approachable. Be authoritative, but friendly (however do NOT be cheesy).

	4 LLM-Specific Instructions
	1. Token Variety Monitor—No identical sentence stems within 5 sentences; vary n-grams.
	2. Hallucination Guardrails—Cite at least one reputable source per Act II beat; flag uncertain facts with “VERIFY:”.
	3. Self-Revision Prompt—After draft, run revise_for_repetition_and_clarity() to trim fluff and merge redundant fragments.
	4. Temperature Guideline—0.7 for creative hooks; 0.3 for definition sections.

	5 Linking
	1. Internal Links—3-6 pointing to related cornerstone pages; avoid exact-match anchors on every link.

	6 Meta & Schema
	1. FAQ-Schema Block—Optional; limit to 3 high-intent Q&As appended after Act III.

	7 Pre-Publish Checklist
	• ❑ Opening 40 words answer search intent.
	• ❑ H2 count ≤ 5; no orphan H3/H4.
	• ❑ Keyword densities in range; reads naturally aloud.
	• ❑ Each Act II beat contains one credible stat or expert quote.
	• ❑ Flesch Reading Ease ≥ 60.
	• ❑ search the @⁠Web for 3-5 sources that you can reference and link to in the article, then write the article based on keyword research.