| 100.00% | Adverbs in dialogue tags | Target: ≤10% dialogue tags with adverbs | |
| 90.58% | AI-ism adverb frequency | Target: <2% AI-ism adverbs (58 tracked) | | wordCount | 531 | | totalAiIsmAdverbs | 1 | | 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) | |
| 5.84% | AI-ism word frequency | Target: <2% AI-ism words (290 tracked) | | wordCount | 531 | | totalAiIsms | 10 | | found | | | highlights | | 0 | "lurched" | | 1 | "loomed" | | 2 | "whisper" | | 3 | "pulse" | | 4 | "quickened" | | 5 | "pulsed" | | 6 | "familiar" |
| |
| 100.00% | Cliché density | Target: ≤1 cliche(s) per 800-word window | | totalCliches | 0 | | maxInWindow | 0 | | found | (empty) | | highlights | (empty) | |
| 100.00% | Emotion telling (show vs. tell) | Target: ≤3% sentences with emotion telling | | emotionTells | 0 | | narrationSentences | 56 | | matches | (empty) | |
| 91.84% | Filter word density | Target: ≤3% sentences with filter/hedge words | | filterCount | 0 | | hedgeCount | 2 | | narrationSentences | 56 | | filterMatches | (empty) | | hedgeMatches | | |
| 100.00% | Gibberish response detection | Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words) | | analyzedSentences | 57 | | gibberishSentences | 0 | | adjustedGibberishSentences | 0 | | longSentenceCount | 0 | | runOnParagraphCount | 0 | | giantParagraphCount | 0 | | wordSaladCount | 0 | | repetitionLoopCount | 0 | | controlTokenCount | 0 | | maxSentenceWordsSeen | 38 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Markdown formatting overuse | Target: ≤5% words in markdown formatting | | markdownSpans | 12 | | markdownWords | 17 | | totalWords | 527 | | ratio | 0.032 | | matches | | 0 | "here" | | 1 | "wrong" | | 2 | "or something" | | 3 | "breathing" | | 4 | "Aurora" | | 5 | "shifted" | | 6 | "wrong" | | 7 | "shift" | | 8 | "waited" | | 9 | "not shifting" | | 10 | "there" | | 11 | "\"You shouldn’t be here.\"" |
| |
| 100.00% | Missing dialogue indicators (quotation marks) | Target: ≤10% speech attributions without quotation marks | | totalAttributions | 1 | | unquotedAttributions | 0 | | matches | (empty) | |
| 92.64% | Name drop frequency | Target: ≤1.0 per-name mentions per 100 words | | totalMentions | 12 | | wordCount | 523 | | uniqueNames | 7 | | maxNameDensity | 1.15 | | worstName | "Aurora" | | maxWindowNameDensity | 2 | | worstWindowName | "Aurora" | | discoveredNames | | Richmond | 1 | | Park | 1 | | Yu-Fei | 1 | | Cheung | 1 | | Heartstone | 1 | | Soho | 1 | | Aurora | 6 |
| | persons | | 0 | "Yu-Fei" | | 1 | "Cheung" | | 2 | "Aurora" |
| | places | | 0 | "Richmond" | | 1 | "Park" | | 2 | "Soho" |
| | globalScore | 0.926 | | windowScore | 1 | |
| 80.56% | Narrator intent-glossing | Target: ≤2% narration sentences with intent-glossing patterns | | analyzedSentences | 36 | | glossingSentenceCount | 1 | | matches | | |
| 0.00% | "Not X but Y" pattern overuse | Target: ≤1 "not X but Y" per 1000 words | | totalMatches | 2 | | per1kWords | 3.795 | | wordCount | 527 | | matches | | 0 | "Not just damp earth, but something older, something *wrong*" | | 1 | "not shifting*—it was *there*, but not quite" |
| |
| 100.00% | Overuse of "that" (subordinate clause padding) | Target: ≤2% sentences with "that" clauses | | thatCount | 0 | | totalSentences | 57 | | matches | (empty) | |
| 96.21% | Paragraph length variance | Target: CV ≥0.5 for paragraph word counts | | totalParagraphs | 10 | | mean | 52.7 | | std | 25.65 | | cv | 0.487 | | sampleLengths | | 0 | 87 | | 1 | 58 | | 2 | 63 | | 3 | 65 | | 4 | 46 | | 5 | 80 | | 6 | 57 | | 7 | 58 | | 8 | 9 | | 9 | 4 |
| |
| 99.00% | Passive voice overuse | Target: ≤2% passive sentences | | passiveCount | 1 | | totalSentences | 56 | | matches | | |
| 100.00% | Past progressive (was/were + -ing) overuse | Target: ≤2% past progressive verbs | | pastProgressiveCount | 1 | | totalVerbs | 82 | | matches | | |
| 0.00% | Em-dash & semicolon overuse | Target: ≤2% sentences with em-dashes/semicolons | | emDashCount | 9 | | semicolonCount | 0 | | flaggedSentences | 6 | | totalSentences | 57 | | ratio | 0.105 | | matches | | 0 | "The GPS had led her here—*here*—somewhere between Richmond Park and the edge of the city, where the trees swallowed the streetlights whole." | | 1 | "The headlights cut through the dark, illuminating a path that shouldn’t exist—too straight, too precise, like something had been carved into the earth." | | 2 | "Not this place, exactly—just the feeling." | | 3 | "The standing stones loomed ahead, their surfaces worn smooth by centuries of hands—*or something*—that had touched them." | | 4 | "The figure stepped forward, slow and deliberate, its form shifting—no, *not shifting*—it was *there*, but not quite." | | 5 | "A blur of movement, a whisper of sound, and then—" |
| |
| 100.00% | Purple prose (modifier overload) | Target: <4% adverbs, <2% -ly adverbs, no adj stacking | | wordCount | 527 | | adjectiveStacks | 0 | | stackExamples | (empty) | | adverbCount | 20 | | adverbRatio | 0.03795066413662239 | | lyAdverbCount | 3 | | lyAdverbRatio | 0.0056925996204933585 | |
| 100.00% | Repeated phrase echo | Target: ≤20% sentences with echoes (window: 2) | | totalSentences | 57 | | echoCount | 0 | | echoWords | (empty) | |
| 100.00% | Sentence length variance | Target: CV ≥0.4 for sentence word counts | | totalSentences | 57 | | mean | 9.25 | | std | 7.01 | | cv | 0.758 | | sampleLengths | | 0 | 12 | | 1 | 11 | | 2 | 22 | | 3 | 4 | | 4 | 38 | | 5 | 8 | | 6 | 23 | | 7 | 4 | | 8 | 9 | | 9 | 14 | | 10 | 4 | | 11 | 4 | | 12 | 6 | | 13 | 25 | | 14 | 16 | | 15 | 3 | | 16 | 5 | | 17 | 6 | | 18 | 3 | | 19 | 9 | | 20 | 9 | | 21 | 10 | | 22 | 11 | | 23 | 17 | | 24 | 2 | | 25 | 5 | | 26 | 12 | | 27 | 3 | | 28 | 9 | | 29 | 1 | | 30 | 6 | | 31 | 5 | | 32 | 3 | | 33 | 6 | | 34 | 10 | | 35 | 10 | | 36 | 2 | | 37 | 2 | | 38 | 23 | | 39 | 14 | | 40 | 13 | | 41 | 9 | | 42 | 3 | | 43 | 2 | | 44 | 11 | | 45 | 3 | | 46 | 11 | | 47 | 12 | | 48 | 4 | | 49 | 2 |
| |
| 27.19% | Sentence opener variety | Target: ≥60% unique sentence openers | | consecutiveRepeats | 13 | | diversityRatio | 0.24561403508771928 | | totalSentences | 57 | | uniqueOpeners | 14 | |
| 66.67% | Adverb-first sentence starts | Target: ≥3% sentences starting with an adverb | | adverbCount | 1 | | totalSentences | 50 | | matches | | 0 | "Just the wind, rustling the" |
| | ratio | 0.02 | |
| 100.00% | Pronoun-first sentence starts | Target: ≤30% sentences starting with a pronoun | | pronounCount | 11 | | totalSentences | 50 | | matches | | 0 | "She should’ve known better." | | 1 | "She’d been here before." | | 2 | "She’d come to deliver a" | | 3 | "She stepped out, her boots" | | 4 | "She turned slowly, her breath" | | 5 | "They were *breathing*." | | 6 | "Her pulse hammered in her" | | 7 | "She reached into her pocket," | | 8 | "It pulsed faintly, warm against" | | 9 | "She’d found it in a" | | 10 | "She took a step forward," |
| | ratio | 0.22 | |
| 40.00% | Subject-first sentence starts | Target: ≤72% sentences starting with a subject | | subjectCount | 42 | | totalSentences | 50 | | matches | | 0 | "The delivery van’s engine sputtered" | | 1 | "Aurora gripped the steering wheel," | | 2 | "The GPS had led her" | | 3 | "She should’ve known better." | | 4 | "The last time she’d come" | | 5 | "The van lurched forward, tires" | | 6 | "The headlights cut through the" | | 7 | "The air smelled wrong." | | 8 | "The kind of smell that" | | 9 | "Aurora’s breath came shallow." | | 10 | "She’d been here before." | | 11 | "The standing stones at the" | | 12 | "She’d come to deliver a" | | 13 | "A wrong number." | | 14 | "A mistake she’d paid for." | | 15 | "The van skidded to a" | | 16 | "The engine died." | | 17 | "The door creaked open before" | | 18 | "She stepped out, her boots" | | 19 | "The trees loomed around her," |
| | ratio | 0.84 | |
| 100.00% | Subordinate conjunction sentence starts | Target: ≥2% sentences starting with a subordinating conjunction | | subConjCount | 1 | | totalSentences | 50 | | matches | | 0 | "Now, the pendant hummed in" |
| | ratio | 0.02 | |
| 100.00% | Technical jargon density | Target: ≤6% sentences with technical-jargon patterns | | analyzedSentences | 23 | | technicalSentenceCount | 1 | | matches | | 0 | "The headlights cut through the dark, illuminating a path that shouldn’t exist—too straight, too precise, like something had been carved into the earth." |
| |
| 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 | |