| 100.00% | Adverbs in dialogue tags | Target: ≤10% dialogue tags with adverbs | |
| 90.14% | AI-ism adverb frequency | Target: <2% AI-ism adverbs (58 tracked) | | wordCount | 1014 | | totalAiIsmAdverbs | 2 | | found | | | highlights | | |
| 100.00% | AI-ism character names | Target: 0 AI-default names (17 tracked, −20% each) | | codexExemptions | (empty) | | found | (empty) | |
| 100.00% | AI-ism location names | Target: 0 AI-default location names (33 tracked, −20% each) | | codexExemptions | (empty) | | found | (empty) | |
| 45.76% | AI-ism word frequency | Target: <2% AI-ism words (290 tracked) | | wordCount | 1014 | | totalAiIsms | 11 | | found | | | highlights | | 0 | "familiar" | | 1 | "beacon" | | 2 | "pumping" | | 3 | "gleaming" | | 4 | "scanned" | | 5 | "chill" | | 6 | "intricate" | | 7 | "etched" | | 8 | "resolved" | | 9 | "potential" | | 10 | "warmth" |
| |
| 100.00% | Cliché density | Target: ≤1 cliche(s) per 800-word window | | totalCliches | 1 | | maxInWindow | 1 | | found | | 0 | | label | "clenched jaw/fists" | | count | 1 |
|
| | highlights | | |
| 100.00% | Emotion telling (show vs. tell) | Target: ≤3% sentences with emotion telling | | emotionTells | 0 | | narrationSentences | 94 | | matches | (empty) | |
| 100.00% | Filter word density | Target: ≤3% sentences with filter/hedge words | | filterCount | 1 | | hedgeCount | 0 | | narrationSentences | 94 | | filterMatches | | | hedgeMatches | (empty) | |
| 100.00% | Gibberish response detection | Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words) | | analyzedSentences | 94 | | gibberishSentences | 0 | | adjustedGibberishSentences | 0 | | longSentenceCount | 0 | | runOnParagraphCount | 0 | | giantParagraphCount | 0 | | wordSaladCount | 0 | | repetitionLoopCount | 0 | | controlTokenCount | 0 | | maxSentenceWordsSeen | 37 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Markdown formatting overuse | Target: ≤5% words in markdown formatting | | markdownSpans | 1 | | markdownWords | 3 | | totalWords | 1003 | | ratio | 0.003 | | matches | | |
| 100.00% | Missing dialogue indicators (quotation marks) | Target: ≤10% speech attributions without quotation marks | | totalAttributions | 2 | | unquotedAttributions | 0 | | matches | (empty) | |
| 100.00% | Name drop frequency | Target: ≤1.0 per-name mentions per 100 words | | totalMentions | 25 | | wordCount | 1003 | | uniqueNames | 12 | | maxNameDensity | 0.7 | | worstName | "Quinn" | | maxWindowNameDensity | 1.5 | | worstWindowName | "Quinn" | | discoveredNames | | London | 1 | | Harlow | 1 | | Quinn | 7 | | Soho | 1 | | Morris | 4 | | Raven | 1 | | Nest | 1 | | Camden | 1 | | Tube | 1 | | Veil | 2 | | Market | 2 | | Protocol | 3 |
| | persons | | 0 | "Harlow" | | 1 | "Quinn" | | 2 | "Morris" | | 3 | "Raven" | | 4 | "Protocol" |
| | places | | 0 | "London" | | 1 | "Soho" | | 2 | "Camden" |
| | globalScore | 1 | | windowScore | 1 | |
| 100.00% | Narrator intent-glossing | Target: ≤2% narration sentences with intent-glossing patterns | | analyzedSentences | 60 | | glossingSentenceCount | 0 | | matches | (empty) | |
| 100.00% | "Not X but Y" pattern overuse | Target: ≤1 "not X but Y" per 1000 words | | totalMatches | 1 | | per1kWords | 0.997 | | wordCount | 1003 | | matches | | 0 | "not twenty-one, but the discipline was hard-wired" |
| |
| 100.00% | Overuse of "that" (subordinate clause padding) | Target: ≤2% sentences with "that" clauses | | thatCount | 0 | | totalSentences | 94 | | matches | (empty) | |
| 100.00% | Paragraph length variance | Target: CV ≥0.5 for paragraph word counts | | totalParagraphs | 19 | | mean | 52.79 | | std | 29.81 | | cv | 0.565 | | sampleLengths | | 0 | 79 | | 1 | 99 | | 2 | 67 | | 3 | 70 | | 4 | 58 | | 5 | 4 | | 6 | 28 | | 7 | 49 | | 8 | 1 | | 9 | 90 | | 10 | 57 | | 11 | 72 | | 12 | 31 | | 13 | 37 | | 14 | 29 | | 15 | 24 | | 16 | 48 | | 17 | 45 | | 18 | 115 |
| |
| 86.60% | Passive voice overuse | Target: ≤2% passive sentences | | passiveCount | 5 | | totalSentences | 94 | | matches | | 0 | "was gone" | | 1 | "was gone" | | 2 | "was supposed" | | 3 | "being opened" | | 4 | "was etched" |
| |
| 100.00% | Past progressive (was/were + -ing) overuse | Target: ≤2% past progressive verbs | | pastProgressiveCount | 0 | | totalVerbs | 152 | | matches | (empty) | |
| 100.00% | Em-dash & semicolon overuse | Target: ≤2% sentences with em-dashes/semicolons | | emDashCount | 0 | | semicolonCount | 0 | | flaggedSentences | 0 | | totalSentences | 94 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Purple prose (modifier overload) | Target: <4% adverbs, <2% -ly adverbs, no adj stacking | | wordCount | 1014 | | adjectiveStacks | 0 | | stackExamples | (empty) | | adverbCount | 24 | | adverbRatio | 0.023668639053254437 | | lyAdverbCount | 8 | | lyAdverbRatio | 0.007889546351084813 | |
| 100.00% | Repeated phrase echo | Target: ≤20% sentences with echoes (window: 2) | | totalSentences | 94 | | echoCount | 0 | | echoWords | (empty) | |
| 100.00% | Sentence length variance | Target: CV ≥0.4 for sentence word counts | | totalSentences | 94 | | mean | 10.67 | | std | 7.56 | | cv | 0.708 | | sampleLengths | | 0 | 16 | | 1 | 22 | | 2 | 15 | | 3 | 26 | | 4 | 20 | | 5 | 3 | | 6 | 15 | | 7 | 34 | | 8 | 13 | | 9 | 4 | | 10 | 10 | | 11 | 12 | | 12 | 17 | | 13 | 11 | | 14 | 2 | | 15 | 2 | | 16 | 11 | | 17 | 12 | | 18 | 23 | | 19 | 7 | | 20 | 12 | | 21 | 6 | | 22 | 10 | | 23 | 12 | | 24 | 17 | | 25 | 11 | | 26 | 6 | | 27 | 15 | | 28 | 9 | | 29 | 4 | | 30 | 4 | | 31 | 4 | | 32 | 20 | | 33 | 8 | | 34 | 4 | | 35 | 21 | | 36 | 9 | | 37 | 3 | | 38 | 3 | | 39 | 1 | | 40 | 1 | | 41 | 5 | | 42 | 11 | | 43 | 29 | | 44 | 15 | | 45 | 4 | | 46 | 15 | | 47 | 11 | | 48 | 13 | | 49 | 8 |
| |
| 36.88% | Sentence opener variety | Target: ≥60% unique sentence openers | | consecutiveRepeats | 17 | | diversityRatio | 0.32978723404255317 | | totalSentences | 94 | | uniqueOpeners | 31 | |
| 100.00% | Adverb-first sentence starts | Target: ≥3% sentences starting with an adverb | | adverbCount | 3 | | totalSentences | 89 | | matches | | 0 | "Then he was gone." | | 1 | "Of course, no signal." | | 2 | "Properly on her own." |
| | ratio | 0.034 | |
| 89.66% | Pronoun-first sentence starts | Target: ≤30% sentences starting with a pronoun | | pronounCount | 29 | | totalSentences | 89 | | matches | | 0 | "Her target, a lean figure" | | 1 | "He moved with an unnatural" | | 2 | "She ignored it." | | 3 | "Her world had narrowed to" | | 4 | "Her sergeant would have her" | | 5 | "Her sergeant wasn’t the one" | | 6 | "He sprinted on, his legs" | | 7 | "She was forty-one, not twenty-one," | | 8 | "She matched his pace, the" | | 9 | "He veered north, towards Camden," | | 10 | "She could taste the city" | | 11 | "She scanned the area, her" | | 12 | "It was a place people" | | 13 | "She saw the faint scuff" | | 14 | "She saw the small, pale" | | 15 | "She approached cautiously, her hand" | | 16 | "It carried the scent of" | | 17 | "It was the smell of" | | 18 | "Her eyes fell on the" | | 19 | "It wasn't a coin." |
| | ratio | 0.326 | |
| 49.89% | Subject-first sentence starts | Target: ≤72% sentences starting with a subject | | subjectCount | 73 | | totalSentences | 89 | | matches | | 0 | "The rain fell in sheets," | | 1 | "Detective Harlow Quinn’s lungs burned," | | 2 | "Her target, a lean figure" | | 3 | "He moved with an unnatural" | | 4 | "Quinn’s worn leather watch, a" | | 5 | "She ignored it." | | 6 | "Her world had narrowed to" | | 7 | "This was the third break-in" | | 8 | "This time was different." | | 9 | "This time, she was close" | | 10 | "The suspect ducked sharply, disappearing" | | 11 | "Quinn followed, her hand instinctively" | | 12 | "Tonight, there was only the" | | 13 | "Her sergeant would have her" | | 14 | "Her sergeant wasn’t the one" | | 15 | "The alley opened onto a" | | 16 | "The suspect didn’t even glance" | | 17 | "He sprinted on, his legs" | | 18 | "Quinn’s own muscles screamed in" | | 19 | "She was forty-one, not twenty-one," |
| | ratio | 0.82 | |
| 100.00% | Subordinate conjunction sentence starts | Target: ≥2% sentences starting with a subordinating conjunction | | subConjCount | 2 | | totalSentences | 89 | | matches | | 0 | "To go down was to" | | 1 | "To stay was to admit" |
| | ratio | 0.022 | |
| 82.07% | Technical jargon density | Target: ≤6% sentences with technical-jargon patterns | | analyzedSentences | 47 | | technicalSentenceCount | 4 | | matches | | 0 | "A double-decker bus groaned past, its red paint gleaming, its windows showing her own distorted reflection: a woman drenched and furious." | | 1 | "Her radio crackled on her shoulder, a burst of static that resolved into nothing." | | 2 | "This was an unknown environment, a potential tactical nightmare with no intel, no support, and an exit that might vanish at any moment." | | 3 | "The people who turned her partner into a closed case file were down there." |
| |
| 100.00% | Useless dialogue additions | Target: ≤5% dialogue tags with trailing filler fragments | | totalTags | 0 | | uselessAdditionCount | 0 | | matches | (empty) | |
| 100.00% | Dialogue tag variety (said vs. fancy) | Target: ≤10% fancy dialogue tags | |