| 100.00% | Adverbs in dialogue tags | Target: ≤10% dialogue tags with adverbs | | totalTags | 14 | | adverbTagCount | 2 | | adverbTags | | 0 | "Eva glanced around [around]" | | 1 | "Eva leaned back [back]" |
| | dialogueSentences | 70 | | tagDensity | 0.2 | | leniency | 0.4 | | rawRatio | 0.143 | | effectiveRatio | 0.057 | |
| 87.43% | AI-ism adverb frequency | Target: <2% AI-ism adverbs (58 tracked) | | wordCount | 1591 | | totalAiIsmAdverbs | 4 | | found | | | highlights | | |
| 100.00% | AI-ism character names | Target: 0 AI-default names (16 tracked, −20% each) | | codexExemptions | | | found | (empty) | |
| 100.00% | AI-ism location names | Target: 0 AI-default location names (33 tracked, −20% each) | | codexExemptions | (empty) | | found | (empty) | |
| 59.15% | AI-ism word frequency | Target: <2% AI-ism words (290 tracked) | | wordCount | 1591 | | totalAiIsms | 13 | | found | | | highlights | | 0 | "shattered" | | 1 | "echoing" | | 2 | "scanned" | | 3 | "flickered" | | 4 | "eyebrow" | | 5 | "weight" | | 6 | "wavered" | | 7 | "absolutely" | | 8 | "glinting" | | 9 | "silk" | | 10 | "unspoken" | | 11 | "trembled" | | 12 | "churned" |
| |
| 100.00% | Cliché density | Target: ≤1 cliche(s) per 800-word window | | totalCliches | 1 | | maxInWindow | 1 | | found | | 0 | | label | "everything happened fast" | | count | 1 |
|
| | highlights | | 0 | "everything happened fast" |
| |
| 100.00% | Emotion telling (show vs. tell) | Target: ≤3% sentences with emotion telling | | emotionTells | 0 | | narrationSentences | 119 | | matches | (empty) | |
| 100.00% | Filter word density | Target: ≤3% sentences with filter/hedge words | | filterCount | 0 | | hedgeCount | 0 | | narrationSentences | 119 | | filterMatches | (empty) | | hedgeMatches | (empty) | |
| 100.00% | Gibberish response detection | Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words) | | analyzedSentences | 175 | | gibberishSentences | 0 | | adjustedGibberishSentences | 0 | | longSentenceCount | 0 | | runOnParagraphCount | 0 | | giantParagraphCount | 0 | | wordSaladCount | 0 | | repetitionLoopCount | 0 | | controlTokenCount | 0 | | maxSentenceWordsSeen | 43 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Markdown formatting overuse | Target: ≤5% words in markdown formatting | | markdownSpans | 1 | | markdownWords | 8 | | totalWords | 1589 | | ratio | 0.005 | | matches | | 0 | "Think about dinner. I really am sorry. E" |
| |
| 100.00% | Missing dialogue indicators (quotation marks) | Target: ≤10% speech attributions without quotation marks | | totalAttributions | 8 | | unquotedAttributions | 0 | | matches | (empty) | |
| 0.00% | Name drop frequency | Target: ≤1.0 per-name mentions per 100 words | | totalMentions | 86 | | wordCount | 1086 | | uniqueNames | 18 | | maxNameDensity | 2.76 | | worstName | "Eva" | | maxWindowNameDensity | 6 | | worstWindowName | "Eva" | | discoveredNames | | Silas | 9 | | Raven | 3 | | Nest | 3 | | Eva | 30 | | Pre-Law | 1 | | London | 3 | | Rory | 19 | | Manchester | 1 | | Burberry | 2 | | Mayfair | 1 | | Guinness | 1 | | French | 2 | | Instagram | 1 | | Friday | 4 | | Soho | 1 | | Translation | 1 | | Mercedes | 1 | | Old | 3 |
| | persons | | 0 | "Silas" | | 1 | "Raven" | | 2 | "Nest" | | 3 | "Eva" | | 4 | "Rory" |
| | places | | 0 | "London" | | 1 | "Manchester" | | 2 | "Friday" | | 3 | "Soho" |
| | globalScore | 0.119 | | windowScore | 0 | |
| 81.51% | Narrator intent-glossing | Target: ≤2% narration sentences with intent-glossing patterns | | analyzedSentences | 73 | | glossingSentenceCount | 2 | | matches | | 0 | "r sell out had apparently found her price" | | 1 | "looked like genuine regret swimming benea" |
| |
| 100.00% | "Not X but Y" pattern overuse | Target: ≤1 "not X but Y" per 1000 words | | totalMatches | 0 | | per1kWords | 0 | | wordCount | 1589 | | matches | (empty) | |
| 100.00% | Overuse of "that" (subordinate clause padding) | Target: ≤2% sentences with "that" clauses | | thatCount | 0 | | totalSentences | 175 | | matches | (empty) | |
| 100.00% | Paragraph length variance | Target: CV ≥0.5 for paragraph word counts | | totalParagraphs | 92 | | mean | 17.27 | | std | 15.25 | | cv | 0.883 | | sampleLengths | | 0 | 8 | | 1 | 63 | | 2 | 45 | | 3 | 5 | | 4 | 1 | | 5 | 39 | | 6 | 17 | | 7 | 45 | | 8 | 3 | | 9 | 18 | | 10 | 4 | | 11 | 27 | | 12 | 1 | | 13 | 32 | | 14 | 28 | | 15 | 15 | | 16 | 2 | | 17 | 11 | | 18 | 14 | | 19 | 5 | | 20 | 30 | | 21 | 5 | | 22 | 30 | | 23 | 1 | | 24 | 33 | | 25 | 35 | | 26 | 2 | | 27 | 1 | | 28 | 2 | | 29 | 15 | | 30 | 43 | | 31 | 1 | | 32 | 55 | | 33 | 11 | | 34 | 4 | | 35 | 6 | | 36 | 1 | | 37 | 29 | | 38 | 4 | | 39 | 25 | | 40 | 3 | | 41 | 6 | | 42 | 4 | | 43 | 18 | | 44 | 39 | | 45 | 7 | | 46 | 20 | | 47 | 7 | | 48 | 42 | | 49 | 23 |
| |
| 100.00% | Passive voice overuse | Target: ≤2% passive sentences | | passiveCount | 1 | | totalSentences | 119 | | matches | | |
| 100.00% | Past progressive (was/were + -ing) overuse | Target: ≤2% past progressive verbs | | pastProgressiveCount | 0 | | totalVerbs | 198 | | matches | (empty) | |
| 100.00% | Em-dash & semicolon overuse | Target: ≤2% sentences with em-dashes/semicolons | | emDashCount | 0 | | semicolonCount | 0 | | flaggedSentences | 0 | | totalSentences | 175 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Purple prose (modifier overload) | Target: <4% adverbs, <2% -ly adverbs, no adj stacking | | wordCount | 1087 | | adjectiveStacks | 0 | | stackExamples | (empty) | | adverbCount | 40 | | adverbRatio | 0.03679852805887764 | | lyAdverbCount | 12 | | lyAdverbRatio | 0.011039558417663294 | |
| 100.00% | Repeated phrase echo | Target: ≤20% sentences with echoes (window: 2) | | totalSentences | 175 | | echoCount | 0 | | echoWords | (empty) | |
| 100.00% | Sentence length variance | Target: CV ≥0.4 for sentence word counts | | totalSentences | 175 | | mean | 9.08 | | std | 6.78 | | cv | 0.746 | | sampleLengths | | 0 | 8 | | 1 | 20 | | 2 | 43 | | 3 | 10 | | 4 | 21 | | 5 | 4 | | 6 | 10 | | 7 | 5 | | 8 | 1 | | 9 | 3 | | 10 | 21 | | 11 | 15 | | 12 | 17 | | 13 | 11 | | 14 | 15 | | 15 | 19 | | 16 | 3 | | 17 | 18 | | 18 | 4 | | 19 | 7 | | 20 | 8 | | 21 | 12 | | 22 | 1 | | 23 | 14 | | 24 | 13 | | 25 | 5 | | 26 | 14 | | 27 | 14 | | 28 | 15 | | 29 | 2 | | 30 | 7 | | 31 | 4 | | 32 | 14 | | 33 | 5 | | 34 | 14 | | 35 | 14 | | 36 | 2 | | 37 | 4 | | 38 | 1 | | 39 | 30 | | 40 | 1 | | 41 | 5 | | 42 | 14 | | 43 | 8 | | 44 | 1 | | 45 | 5 | | 46 | 7 | | 47 | 15 | | 48 | 13 | | 49 | 2 |
| |
| 68.00% | Sentence opener variety | Target: ≥60% unique sentence openers | | consecutiveRepeats | 6 | | diversityRatio | 0.42857142857142855 | | totalSentences | 175 | | uniqueOpeners | 75 | |
| 61.16% | Adverb-first sentence starts | Target: ≥3% sentences starting with an adverb | | adverbCount | 2 | | totalSentences | 109 | | matches | | 0 | "Then she deleted the contact," | | 1 | "Just another ghost." |
| | ratio | 0.018 | |
| 100.00% | Pronoun-first sentence starts | Target: ≤30% sentences starting with a pronoun | | pronounCount | 22 | | totalSentences | 109 | | matches | | 0 | "She dumped the glass into" | | 1 | "Her coat was Burberry, her" | | 2 | "Their eyes met." | | 3 | "His eyebrow rose fractionally, a" | | 4 | "She stepped from behind the" | | 5 | "Her perfume arrived first, something" | | 6 | "She pulled back, hands still" | | 7 | "His attention remained fixed on" | | 8 | "She shed her coat, revealing" | | 9 | "They claimed a corner table," | | 10 | "She sipped her champagne" | | 11 | "Her eyes held something then," | | 12 | "It felt obscenely heavy." | | 13 | "She left three twenties on" | | 14 | "She left trailing expensive perfume" | | 15 | "His knee injury made him" | | 16 | "She pulled pints and collected" | | 17 | "Her phone buzzed." | | 18 | "I really am sorry." | | 19 | "His expression asked a silent" |
| | ratio | 0.202 | |
| 28.81% | Subject-first sentence starts | Target: ≤72% sentences starting with a subject | | subjectCount | 94 | | totalSentences | 109 | | matches | | 0 | "The pint glass shattered against" | | 1 | "Rory's hands moved on autopilot," | | 2 | "Friday nights at The Raven's" | | 3 | "She dumped the glass into" | | 4 | "Movement caught her eye near" | | 5 | "The woman moved like she" | | 6 | "Recognition hit like cold water." | | 7 | "Cardiff flooded back: Eva's laughter" | | 8 | "Eva who'd convinced her London" | | 9 | "Eva who'd vanished three weeks" | | 10 | "The woman at the door" | | 11 | "Her coat was Burberry, her" | | 12 | "The girl who'd worn charity" | | 13 | "Their eyes met." | | 14 | "Eva's expression flickered through surprise," | | 15 | "Silas glanced up from pulling" | | 16 | "His eyebrow rose fractionally, a" | | 17 | "She stepped from behind the" | | 18 | "Eva crossed the distance between" | | 19 | "Her perfume arrived first, something" |
| | ratio | 0.862 | |
| 0.00% | Subordinate conjunction sentence starts | Target: ≥2% sentences starting with a subordinating conjunction | | subConjCount | 0 | | totalSentences | 109 | | matches | (empty) | | ratio | 0 | |
| 9.97% | Technical jargon density | Target: ≤6% sentences with technical-jargon patterns | | analyzedSentences | 43 | | technicalSentenceCount | 8 | | matches | | 0 | "Friday nights at The Raven's Nest carried their own rhythm: laughter swelling near the dartboard, the jukebox cycling through decades of forgettable rock, the s…" | | 1 | "Eva who'd convinced her London was salvation, who'd promised a spare room and fresh starts." | | 2 | "Eva who'd vanished three weeks after Rory arrived, leaving only a note about unexpected opportunities in Manchester." | | 3 | "Her coat was Burberry, her hair professionally styled in a way that screamed Mayfair salon." | | 4 | "The girl who'd worn charity shop band tees and sworn she'd never sell out had apparently found her price." | | 5 | "Eva's expression flickered through surprise, calculation, and something that might have been guilt before settling on manufactured delight." | | 6 | "She shed her coat, revealing a silk blouse that cost more than Rory's monthly rent." | | 7 | "Her eyes held something then, something that looked like genuine regret swimming beneath the professional veneer." |
| |
| 100.00% | Useless dialogue additions | Target: ≤5% dialogue tags with trailing filler fragments | | totalTags | 14 | | uselessAdditionCount | 0 | | matches | (empty) | |
| 100.00% | Dialogue tag variety (said vs. fancy) | Target: ≤10% fancy dialogue tags | | totalTags | 2 | | fancyCount | 1 | | fancyTags | | | dialogueSentences | 70 | | tagDensity | 0.029 | | leniency | 0.057 | | rawRatio | 0.5 | | effectiveRatio | 0.029 | |