| 100.00% | Adverbs in dialogue tags | Target: ≤10% dialogue tags with adverbs | | totalTags | 2 | | adverbTagCount | 0 | | adverbTags | (empty) | | dialogueSentences | 8 | | tagDensity | 0.25 | | leniency | 0.5 | | rawRatio | 0 | | effectiveRatio | 0 | |
| 100.00% | AI-ism adverb frequency | Target: <2% AI-ism adverbs (58 tracked) | | wordCount | 1063 | | totalAiIsmAdverbs | 0 | | found | (empty) | | highlights | (empty) | |
| 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) | |
| 67.07% | AI-ism word frequency | Target: <2% AI-ism words (290 tracked) | | wordCount | 1063 | | totalAiIsms | 7 | | found | | | highlights | | 0 | "churned" | | 1 | "glistening" | | 2 | "scanning" | | 3 | "weight" | | 4 | "standard" | | 5 | "remarkable" | | 6 | "absolutely" |
| |
| 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 | 95 | | matches | (empty) | |
| 100.00% | Filter word density | Target: ≤3% sentences with filter/hedge words | | filterCount | 0 | | hedgeCount | 1 | | narrationSentences | 95 | | filterMatches | (empty) | | hedgeMatches | | |
| 100.00% | Gibberish response detection | Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words) | | analyzedSentences | 100 | | 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 | 0 | | markdownWords | 0 | | totalWords | 1061 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Missing dialogue indicators (quotation marks) | Target: ≤10% speech attributions without quotation marks | | totalAttributions | 1 | | unquotedAttributions | 0 | | matches | (empty) | |
| 100.00% | Name drop frequency | Target: ≤1.0 per-name mentions per 100 words | | totalMentions | 26 | | wordCount | 1008 | | uniqueNames | 16 | | maxNameDensity | 0.6 | | worstName | "Quinn" | | maxWindowNameDensity | 1.5 | | worstWindowName | "Quinn" | | discoveredNames | | Raven | 1 | | Nest | 1 | | Wardour | 1 | | Street | 2 | | Saturday | 1 | | Soho | 1 | | Berwick | 1 | | Tube | 2 | | Underground | 1 | | Morris | 2 | | London | 1 | | Quinn | 6 | | Spanish | 1 | | Herrera | 2 | | Seville | 1 | | Looked | 2 |
| | persons | | 0 | "Morris" | | 1 | "Quinn" | | 2 | "Herrera" | | 3 | "Looked" |
| | places | | 0 | "Raven" | | 1 | "Wardour" | | 2 | "Street" | | 3 | "Soho" | | 4 | "Berwick" | | 5 | "London" | | 6 | "Seville" |
| | globalScore | 1 | | windowScore | 1 | |
| 100.00% | Narrator intent-glossing | Target: ≤2% narration sentences with intent-glossing patterns | | analyzedSentences | 63 | | glossingSentenceCount | 1 | | matches | | 0 | "looked like the entrance to an abandoned" |
| |
| 100.00% | "Not X but Y" pattern overuse | Target: ≤1 "not X but Y" per 1000 words | | totalMatches | 1 | | per1kWords | 0.943 | | wordCount | 1061 | | matches | | 0 | "not of guilt but of desperate calculation" |
| |
| 100.00% | Overuse of "that" (subordinate clause padding) | Target: ≤2% sentences with "that" clauses | | thatCount | 1 | | totalSentences | 100 | | matches | | |
| 100.00% | Paragraph length variance | Target: CV ≥0.5 for paragraph word counts | | totalParagraphs | 40 | | mean | 26.53 | | std | 24.7 | | cv | 0.931 | | sampleLengths | | 0 | 7 | | 1 | 48 | | 2 | 7 | | 3 | 5 | | 4 | 2 | | 5 | 59 | | 6 | 64 | | 7 | 7 | | 8 | 8 | | 9 | 22 | | 10 | 13 | | 11 | 62 | | 12 | 2 | | 13 | 4 | | 14 | 77 | | 15 | 11 | | 16 | 38 | | 17 | 17 | | 18 | 12 | | 19 | 31 | | 20 | 3 | | 21 | 61 | | 22 | 11 | | 23 | 94 | | 24 | 8 | | 25 | 67 | | 26 | 19 | | 27 | 36 | | 28 | 54 | | 29 | 37 | | 30 | 5 | | 31 | 9 | | 32 | 3 | | 33 | 49 | | 34 | 19 | | 35 | 11 | | 36 | 3 | | 37 | 49 | | 38 | 19 | | 39 | 8 |
| |
| 90.49% | Passive voice overuse | Target: ≤2% passive sentences | | passiveCount | 4 | | totalSentences | 95 | | matches | | 0 | "been resurrected" | | 1 | "been overtaken" | | 2 | "were gone" | | 3 | "been nailed" |
| |
| 87.01% | Past progressive (was/were + -ing) overuse | Target: ≤2% past progressive verbs | | pastProgressiveCount | 3 | | totalVerbs | 177 | | matches | | 0 | "were being" | | 1 | "wasn't looking" | | 2 | "was watching" |
| |
| 100.00% | Em-dash & semicolon overuse | Target: ≤2% sentences with em-dashes/semicolons | | emDashCount | 0 | | semicolonCount | 0 | | flaggedSentences | 0 | | totalSentences | 100 | | ratio | 0 | | matches | (empty) | |
| 100.00% | Purple prose (modifier overload) | Target: <4% adverbs, <2% -ly adverbs, no adj stacking | | wordCount | 1009 | | adjectiveStacks | 0 | | stackExamples | (empty) | | adverbCount | 24 | | adverbRatio | 0.023785926660059464 | | lyAdverbCount | 11 | | lyAdverbRatio | 0.010901883052527254 | |
| 100.00% | Repeated phrase echo | Target: ≤20% sentences with echoes (window: 2) | | totalSentences | 100 | | echoCount | 0 | | echoWords | (empty) | |
| 100.00% | Sentence length variance | Target: CV ≥0.4 for sentence word counts | | totalSentences | 100 | | mean | 10.61 | | std | 8.89 | | cv | 0.838 | | sampleLengths | | 0 | 7 | | 1 | 31 | | 2 | 2 | | 3 | 2 | | 4 | 13 | | 5 | 7 | | 6 | 5 | | 7 | 2 | | 8 | 16 | | 9 | 5 | | 10 | 21 | | 11 | 7 | | 12 | 2 | | 13 | 8 | | 14 | 18 | | 15 | 6 | | 16 | 25 | | 17 | 7 | | 18 | 8 | | 19 | 7 | | 20 | 8 | | 21 | 22 | | 22 | 1 | | 23 | 3 | | 24 | 9 | | 25 | 22 | | 26 | 1 | | 27 | 7 | | 28 | 9 | | 29 | 23 | | 30 | 2 | | 31 | 4 | | 32 | 5 | | 33 | 21 | | 34 | 20 | | 35 | 17 | | 36 | 14 | | 37 | 11 | | 38 | 18 | | 39 | 4 | | 40 | 2 | | 41 | 1 | | 42 | 13 | | 43 | 6 | | 44 | 7 | | 45 | 4 | | 46 | 5 | | 47 | 2 | | 48 | 3 | | 49 | 2 |
| |
| 71.00% | Sentence opener variety | Target: ≥60% unique sentence openers | | consecutiveRepeats | 9 | | diversityRatio | 0.48 | | totalSentences | 100 | | uniqueOpeners | 48 | |
| 82.30% | Adverb-first sentence starts | Target: ≥3% sentences starting with an adverb | | adverbCount | 2 | | totalSentences | 81 | | matches | | 0 | "Dark curly hair plastered to" | | 1 | "Then, twenty metres ahead, a" |
| | ratio | 0.025 | |
| 61.98% | Pronoun-first sentence starts | Target: ≤30% sentences starting with a pronoun | | pronounCount | 32 | | totalSentences | 81 | | matches | | 0 | "She'd been watching the entrance" | | 1 | "She crossed the street at" | | 2 | "He was fast, she'd give" | | 3 | "He cut left onto Berwick" | | 4 | "Her boots cracked against wet" | | 5 | "She nearly caught him at" | | 6 | "Her fingers grazed the back" | | 7 | "He twisted away and vaulted" | | 8 | "She shook the radio and" | | 9 | "She'd seen that face before," | | 10 | "Her breath came in clouds." | | 11 | "She turned in a slow" | | 12 | "Her hand moved to her" | | 13 | "She pulled out her phone." | | 14 | "She radioed again." | | 15 | "She stood there for three" | | 16 | "She went down." | | 17 | "It had been overtaken." | | 18 | "She was the only one" | | 19 | "His customer examined one between" |
| | ratio | 0.395 | |
| 46.42% | Subject-first sentence starts | Target: ≤72% sentences starting with a subject | | subjectCount | 67 | | totalSentences | 81 | | matches | | 0 | "Quinn spotted him the moment" | | 1 | "She'd been watching the entrance" | | 2 | "She crossed the street at" | | 3 | "The alley spat them both" | | 4 | "A hen party scattered, shrieking." | | 5 | "Quinn ducked past a doorman" | | 6 | "He was fast, she'd give" | | 7 | "He cut left onto Berwick" | | 8 | "Her boots cracked against wet" | | 9 | "Rain had been falling for" | | 10 | "She nearly caught him at" | | 11 | "Her fingers grazed the back" | | 12 | "He twisted away and vaulted" | | 13 | "Quinn followed, breathing hard now," | | 14 | "A broken response." | | 15 | "She shook the radio and" | | 16 | "The hood came down as" | | 17 | "An expression not of guilt" | | 18 | "She'd seen that face before," | | 19 | "The street was empty." |
| | ratio | 0.827 | |
| 0.00% | Subordinate conjunction sentence starts | Target: ≥2% sentences starting with a subordinating conjunction | | subConjCount | 0 | | totalSentences | 81 | | matches | (empty) | | ratio | 0 | |
| 0.00% | Technical jargon density | Target: ≤6% sentences with technical-jargon patterns | | analyzedSentences | 35 | | technicalSentenceCount | 7 | | matches | | 0 | "Moving with the particular urgency of someone who'd clocked they were being watched." | | 1 | "The staircase was old London brick, the kind that sweated regardless of the season." | | 2 | "Emergency lighting, cheap and amber, lined the ceiling at intervals, and around the first bend the sound of the city above vanished entirely, replaced by a low,…" | | 3 | "The vaulted ceiling disappeared into darkness above, the original tile-work still visible in patches between the things that had been nailed and strung and buil…" | | 4 | "She found him at the far end of the platform, in a curtained alcove, arguing in low, rapid Spanish with a man she recognised now with a certainty that landed li…" | | 5 | "He had a bag open on the table, medical, she could see the shape of it, and inside, between rolled bandages and blister packs, sat something that caught the lam…" | | 6 | "Looked at the curtained walls around them and understood, with the particular sick clarity that had been building since Morris died three years ago, that she wa…" |
| |
| 100.00% | Useless dialogue additions | Target: ≤5% dialogue tags with trailing filler fragments | | totalTags | 2 | | uselessAdditionCount | 0 | | matches | (empty) | |
| 100.00% | Dialogue tag variety (said vs. fancy) | Target: ≤10% fancy dialogue tags | | totalTags | 1 | | fancyCount | 0 | | fancyTags | (empty) | | dialogueSentences | 8 | | tagDensity | 0.125 | | leniency | 0.25 | | rawRatio | 0 | | effectiveRatio | 0 | |