| 100.00% | Adverbs in dialogue tags | Target: ≤10% dialogue tags with adverbs | | totalTags | 12 | | adverbTagCount | 1 | | adverbTags | | 0 | "she said slowly [slowly]" |
| | dialogueSentences | 38 | | tagDensity | 0.316 | | leniency | 0.632 | | rawRatio | 0.083 | | effectiveRatio | 0.053 | |
| 85.92% | AI-ism adverb frequency | Target: <2% AI-ism adverbs (58 tracked) | | wordCount | 1065 | | totalAiIsmAdverbs | 3 | | found | | | highlights | | 0 | "tightly" | | 1 | "slightly" | | 2 | "slowly" |
| |
| 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) | |
| 57.75% | AI-ism word frequency | Target: <2% AI-ism words (290 tracked) | | wordCount | 1065 | | totalAiIsms | 9 | | found | | | highlights | | 0 | "measured" | | 1 | "weight" | | 2 | "fractured" | | 3 | "fragmented" | | 4 | "trembled" | | 5 | "aftermath" | | 6 | "scanning" | | 7 | "etched" | | 8 | "silence" |
| |
| 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 | 1 | | narrationSentences | 99 | | matches | | |
| 100.00% | Filter word density | Target: ≤3% sentences with filter/hedge words | | filterCount | 1 | | hedgeCount | 1 | | narrationSentences | 99 | | filterMatches | | | hedgeMatches | | |
| 100.00% | Gibberish response detection | Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words) | | analyzedSentences | 125 | | gibberishSentences | 0 | | adjustedGibberishSentences | 0 | | longSentenceCount | 0 | | runOnParagraphCount | 0 | | giantParagraphCount | 0 | | wordSaladCount | 0 | | repetitionLoopCount | 0 | | controlTokenCount | 0 | | maxSentenceWordsSeen | 31 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Markdown formatting overuse | Target: ≤5% words in markdown formatting | | markdownSpans | 0 | | markdownWords | 0 | | totalWords | 1052 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Missing dialogue indicators (quotation marks) | Target: ≤10% speech attributions without quotation marks | | totalAttributions | 17 | | unquotedAttributions | 0 | | matches | (empty) | |
| 46.77% | Name drop frequency | Target: ≤1.0 per-name mentions per 100 words | | totalMentions | 39 | | wordCount | 775 | | uniqueNames | 11 | | maxNameDensity | 2.06 | | worstName | "Quinn" | | maxWindowNameDensity | 3.5 | | worstWindowName | "Eva" | | discoveredNames | | Harlow | 1 | | Quinn | 16 | | Camden | 1 | | Underground | 1 | | Lorne | 1 | | Pryce | 3 | | Kowalski | 1 | | Eva | 12 | | Hermetic | 1 | | Prague | 1 | | Thames | 1 |
| | persons | | 0 | "Harlow" | | 1 | "Quinn" | | 2 | "Pryce" | | 3 | "Kowalski" | | 4 | "Eva" |
| | places | | | globalScore | 0.468 | | windowScore | 0.5 | |
| 100.00% | Narrator intent-glossing | Target: ≤2% narration sentences with intent-glossing patterns | | analyzedSentences | 47 | | glossingSentenceCount | 0 | | matches | (empty) | |
| 100.00% | "Not X but Y" pattern overuse | Target: ≤1 "not X but Y" per 1000 words | | totalMatches | 1 | | per1kWords | 0.951 | | wordCount | 1052 | | matches | | 0 | "not north, but toward the center of the platform" |
| |
| 100.00% | Overuse of "that" (subordinate clause padding) | Target: ≤2% sentences with "that" clauses | | thatCount | 0 | | totalSentences | 125 | | matches | (empty) | |
| 100.00% | Paragraph length variance | Target: CV ≥0.5 for paragraph word counts | | totalParagraphs | 49 | | mean | 21.47 | | std | 20.16 | | cv | 0.939 | | sampleLengths | | 0 | 19 | | 1 | 82 | | 2 | 78 | | 3 | 6 | | 4 | 2 | | 5 | 51 | | 6 | 17 | | 7 | 33 | | 8 | 19 | | 9 | 20 | | 10 | 7 | | 11 | 4 | | 12 | 5 | | 13 | 21 | | 14 | 36 | | 15 | 13 | | 16 | 82 | | 17 | 4 | | 18 | 26 | | 19 | 4 | | 20 | 16 | | 21 | 3 | | 22 | 15 | | 23 | 4 | | 24 | 43 | | 25 | 18 | | 26 | 14 | | 27 | 19 | | 28 | 41 | | 29 | 3 | | 30 | 10 | | 31 | 6 | | 32 | 1 | | 33 | 31 | | 34 | 27 | | 35 | 24 | | 36 | 7 | | 37 | 17 | | 38 | 19 | | 39 | 15 | | 40 | 7 | | 41 | 59 | | 42 | 2 | | 43 | 19 | | 44 | 18 | | 45 | 46 | | 46 | 17 | | 47 | 13 | | 48 | 9 |
| |
| 91.09% | Passive voice overuse | Target: ≤2% passive sentences | | passiveCount | 4 | | totalSentences | 99 | | matches | | 0 | "was mottled" | | 1 | "was packed" | | 2 | "was closed" | | 3 | "been confiscated" |
| |
| 91.60% | Past progressive (was/were + -ing) overuse | Target: ≤2% past progressive verbs | | pastProgressiveCount | 2 | | totalVerbs | 123 | | matches | | 0 | "was, standing" | | 1 | "was counting" |
| |
| 0.00% | Em-dash & semicolon overuse | Target: ≤2% sentences with em-dashes/semicolons | | emDashCount | 12 | | semicolonCount | 0 | | flaggedSentences | 10 | | totalSentences | 125 | | ratio | 0.08 | | matches | | 0 | "The air tasted of rust and damp concrete, thick with the residue of forgotten commuters and something sharper—something wrong." | | 1 | "She’d found it on the victim—an art dealer named Lorne Pryce—clutched so tightly in his right hand that the coroner had to pry open the fingers at the mortuary." | | 2 | "A figure emerged from behind a collapsed service kiosk—short, wiry, hunched slightly under the weight of a battered leather satchel." | | 3 | "But the skin on his neck was mottled with a web of violet lines—like bruising, but too symmetrical, too deliberate." | | 4 | "A faint shimmer in the air above his outstretched hand—like heat haze on tarmac, but cold to the touch when she passed her palm over it." | | 5 | "The needle spun wildly, then locked—not north, but toward the center of the platform, where the air seemed to dip, like an invisible funnel." | | 6 | "Pryce had been an art dealer, yes—but his last acquisition had been a 15th-century grimoire linked to banned Hermetic sects." | | 7 | "The dirt was packed down, but near the rails, something glinted—tiny, metallic." | | 8 | "Morris—her partner—had gone in first." | | 9 | "But now she saw something the coroner had missed—a thin groove carved into the inside of his ring finger." |
| |
| 100.00% | Purple prose (modifier overload) | Target: <4% adverbs, <2% -ly adverbs, no adj stacking | | wordCount | 786 | | adjectiveStacks | 0 | | stackExamples | (empty) | | adverbCount | 30 | | adverbRatio | 0.03816793893129771 | | lyAdverbCount | 9 | | lyAdverbRatio | 0.011450381679389313 | |
| 100.00% | Repeated phrase echo | Target: ≤20% sentences with echoes (window: 2) | | totalSentences | 125 | | echoCount | 0 | | echoWords | (empty) | |
| 100.00% | Sentence length variance | Target: CV ≥0.4 for sentence word counts | | totalSentences | 125 | | mean | 8.42 | | std | 6.86 | | cv | 0.816 | | sampleLengths | | 0 | 19 | | 1 | 24 | | 2 | 30 | | 3 | 11 | | 4 | 6 | | 5 | 11 | | 6 | 29 | | 7 | 22 | | 8 | 27 | | 9 | 6 | | 10 | 2 | | 11 | 10 | | 12 | 2 | | 13 | 20 | | 14 | 8 | | 15 | 11 | | 16 | 2 | | 17 | 2 | | 18 | 2 | | 19 | 11 | | 20 | 14 | | 21 | 4 | | 22 | 2 | | 23 | 13 | | 24 | 15 | | 25 | 4 | | 26 | 2 | | 27 | 18 | | 28 | 7 | | 29 | 4 | | 30 | 5 | | 31 | 16 | | 32 | 5 | | 33 | 3 | | 34 | 14 | | 35 | 5 | | 36 | 8 | | 37 | 6 | | 38 | 10 | | 39 | 3 | | 40 | 21 | | 41 | 2 | | 42 | 3 | | 43 | 7 | | 44 | 20 | | 45 | 10 | | 46 | 4 | | 47 | 5 | | 48 | 10 | | 49 | 4 |
| |
| 66.13% | Sentence opener variety | Target: ≥60% unique sentence openers | | consecutiveRepeats | 4 | | diversityRatio | 0.416 | | totalSentences | 125 | | uniqueOpeners | 52 | |
| 100.00% | Adverb-first sentence starts | Target: ≥3% sentences starting with an adverb | | adverbCount | 5 | | totalSentences | 75 | | matches | | 0 | "More like she was counting" | | 1 | "Then she saw it." | | 2 | "Then reached into her satchel." | | 3 | "Supposedly destroyed in transit." | | 4 | "Just those same violet tracings" |
| | ratio | 0.067 | |
| 100.00% | Pronoun-first sentence starts | Target: ≤30% sentences starting with a pronoun | | pronounCount | 11 | | totalSentences | 75 | | matches | | 0 | "Her breath came steady, but" | | 1 | "She’d found it on the" | | 2 | "She didn’t look nervous." | | 3 | "She stepped forward, her leather" | | 4 | "She’d seen the file." | | 5 | "It hadn’t been." | | 6 | "she said slowly" | | 7 | "She dropped to one knee," | | 8 | "She brushed it clear." | | 9 | "She pulled out her notepad," | | 10 | "They found his body slumped" |
| | ratio | 0.147 | |
| 100.00% | Subject-first sentence starts | Target: ≤72% sentences starting with a subject | | subjectCount | 44 | | totalSentences | 75 | | matches | | 0 | "The air tasted of rust" | | 1 | "Detective Harlow Quinn descended the" | | 2 | "The beam of her torch" | | 3 | "Her breath came steady, but" | | 4 | "This wasn’t just another crime" | | 5 | "She’d found it on the" | | 6 | "The voice came from the" | | 7 | "A figure emerged from behind" | | 8 | "Eva said, tucking a loose" | | 9 | "She didn’t look nervous." | | 10 | "Quinn replied, sweeping the torch" | | 11 | "Quinn studied her." | | 12 | "She stepped forward, her leather" | | 13 | "Pryce lay facedown at the" | | 14 | "The pathologist’s preliminary report said" | | 15 | "Quinn knelt, gloved fingers brushing" | | 16 | "The fabric was dry." | | 17 | "The floor beneath him, too." | | 18 | "A faint shimmer in the" | | 19 | "Eva didn’t answer right away." |
| | ratio | 0.587 | |
| 66.67% | Subordinate conjunction sentence starts | Target: ≥2% sentences starting with a subordinating conjunction | | subConjCount | 1 | | totalSentences | 75 | | matches | | 0 | "And, until three nights ago," |
| | ratio | 0.013 | |
| 100.00% | Technical jargon density | Target: ≤6% sentences with technical-jargon patterns | | analyzedSentences | 28 | | technicalSentenceCount | 1 | | matches | | 0 | "And now here she was, standing in a sealed-off substation that hadn’t seen human traffic in over thirty years, the official records showing its decommission in …" |
| |
| 100.00% | Useless dialogue additions | Target: ≤5% dialogue tags with trailing filler fragments | | totalTags | 12 | | uselessAdditionCount | 0 | | matches | (empty) | |
| 100.00% | Dialogue tag variety (said vs. fancy) | Target: ≤10% fancy dialogue tags | | totalTags | 9 | | fancyCount | 1 | | fancyTags | | | dialogueSentences | 38 | | tagDensity | 0.237 | | leniency | 0.474 | | rawRatio | 0.111 | | effectiveRatio | 0.053 | |