@houtini/voice-analyser 1.3.1 → 1.4.0

package/README.md

@@ -1,19 +1,18 @@
-# Voice Analyser
+# Voice Analyser MCP
 
 [![npm version](https://img.shields.io/npm/v/@houtini/voice-analyser)](https://www.npmjs.com/package/@houtini/voice-analyser)
+[![Known Vulnerabilities](https://snyk.io/test/github/houtini-ai/voice-analyser-mcp/badge.svg)](https://snyk.io/test/github/houtini-ai/voice-analyser-mcp)
 [![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
 
-> **Experimental library** for extracting statistical voice models from your published writing. Generates immersive style guides that teach LLMs to replicate how you actually write - not through rules, but through examples and rhythm patterns.
+> MCP server that analyses your published writing and generates executable style guides for voice-matched content creation.
 
-## Why This Exists
+## What This Does
 
-Traditional style guides list rules: "Use short sentences. Vary paragraph length. Include personal anecdotes."
+I built this because traditional style guides don't work. They tell you "use short sentences" and "vary paragraph length" - rules that sound helpful but produce robotic output when you try to follow them.
 
-This doesn't work. Writers don't follow rules - they channel voice.
+This tool extracts statistical patterns from your published writing and generates a style guide that teaches through **zero tolerance rules, phrase libraries, and validation checklists** rather than vague principles.
 
-This tool extracts the statistical fingerprint of *your writing* and presents it as immersive examples with annotations showing *what makes each passage feel human*. The goal is voice replication through pattern recognition, not rule compliance.
-
-**Status:** Experimental. The approach works but is under active development.
+Version 1.4.0 focuses on executable instructions: forbidden word lists with alternatives, 50+ actual phrases from your corpus, and checkbox validation that catches AI slop before you publish.
 
 ## Installation
 
@@ -32,217 +31,191 @@ Add to your `claude_desktop_config.json`:
 }
 ```
 
-**Config locations:**
+**Config file locations:**
 - Windows: `%APPDATA%\Claude\claude_desktop_config.json`
 - macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
 - Linux: `~/.config/Claude/claude_desktop_config.json`
 
 Restart Claude Desktop after saving.
 
-### Requirements
-
-- Node.js 20+
+**Requirements:** Node.js 20+
 
 ## Quick Start
 
-### 1. Create an Output Directory
+### 1. Create Output Directory
 
-First, create a directory where your corpus and analysis will be stored:
+Pick a directory for corpus storage and analysis:
 
 ```
 C:\writing\voice-models\        (Windows)
 ~/writing/voice-models/          (Mac/Linux)
 ```
 
-This directory will contain:
-- Collected articles (markdown)
-- Analysis JSON files
-- Generated voice guides
+This holds collected articles, analysis JSON files, and generated style guides.
 
-### 2. Collect Your Writing
+### 2. Collect Writing Corpus
 
+In Claude Desktop:
 ```
 Collect corpus from https://yoursite.com/post-sitemap.xml 
 Save as "my-voice" in "C:\writing\voice-models"
 ```
 
-The tool needs:
-- `sitemap_url` - Your XML sitemap URL
-- `output_name` - A name for this corpus (e.g., "my-voice", "blog-posts")
-- `output_dir` - The directory you created above
+Parameters:
+- `sitemap_url` - XML sitemap URL
+- `output_name` - Corpus identifier (e.g., "my-voice")
+- `output_dir` - Directory you created above
+- `max_articles` - Optional limit (default: 100)
 
-**Example with all parameters:**
-```
-Collect corpus from https://example.com/post-sitemap.xml
-Output name: "technical-writing"  
-Output directory: "C:\writing\voice-models"
-Maximum articles: 50
-```
+The tool crawls your sitemap, extracts clean content, and saves markdown files.
 
-### 3. Analyse Patterns
+### 3. Run Analysis
 
 ```
 Analyse corpus "my-voice" in directory "C:\writing\voice-models"
 ```
 
-This runs 14 analysers covering vocabulary, sentence structure, voice markers, argument flow, and paragraph transitions.
+This runs 15+ analysers covering:
+- Vocabulary tiers (AI slop detection, formality scoring)
+- Phrase extraction (opening patterns, transitions, caveats)
+- Sentence structure and rhythm
+- Voice markers and conversational devices
+- Punctuation habits
 
-### 4. Generate Voice Guide
+### 4. Generate Style Guide v4
 
 ```
-Generate narrative guide for "my-voice" in directory "C:\writing\voice-models"
+Generate style guide for "my-voice" in directory "C:\writing\voice-models"
 ```
 
-Creates an immersive style guide with annotated examples at:
-`C:\writing\voice-models\my-voice\writing_style_my-voice_narrative.md`
+Creates an example-first guide at:
+`C:\writing\voice-models\my-voice\writing_style_my-voice.md`
 
-## Using the Voice Guide
+## What v1.4.0 Changed
 
-Once generated, the voice guide can be loaded into any LLM conversation to help it write in your voice.
+Previous versions generated statistical analysis that was accurate but not useful for writing. v1.4.0 restructures the output:
 
-### Loading the Guide
+**Before:** 60% statistics, 40% guidance  
+**After:** 70% examples, 30% statistics
 
-```
-Load the file C:\writing\voice-models\my-voice\writing_style_my-voice_narrative.md 
-and use it as a reference for all writing in this conversation.
-```
+### New Style Guide Structure
 
-### Example Prompts for Content Generation
+**Part 1: Zero Tolerance Rules**
+- Forbidden vocabulary (AI slop) with alternatives
+- Formal words flagged with casual replacements
+- Punctuation rules (em-dash detection)
 
-**Blog post:**
-```
-Using the voice guide as your reference, write a blog post about [topic].
-
-Key requirements:
-- Match the sentence rhythm patterns shown in the examples
-- Use the conversational devices naturally (not forced)
-- Include the micro-rhythms: mid-thought pivots, embedded uncertainty, present-tense immediacy
-- Vary sentence length as shown in the statistical targets
-- Use British/American spelling as indicated in the guide
-```
+**Part 2: Phrase Library (50+ Examples)**
+- Opening patterns (personal story, direct action, protective warnings)
+- Equipment references (possessive vs generic)
+- Caveat phrases (honesty markers)
+- Transition patterns
 
-**Technical article:**
-```
-Reference the voice guide and write a technical explanation of [concept].
-
-Channel the voice by:
-- Opening with the pattern shown in "Opening Moves" section
-- Using specific product/tool names, not generic references  
-- Including admissions of complexity or uncertainty where authentic
-- Following the argument flow patterns from the guide
-- Matching the punctuation habits (especially dash usage)
-```
+**Part 3: Sentence Patterns**
+- Rhythm variation targets with corpus examples
+- First-person usage frequency
+- Natural sentence flow demonstrations
+
+**Part 4: Validation Checklist**
+- Critical rules (must pass)
+- Voice match rules (should pass)
+- Actionable checkbox format
+
+**Part 5: Quick Reference**
+- Top phrases by frequency
+- Statistics summary
+
+## Using the Style Guide
+
+Load the generated guide into Claude conversations:
 
-**Product review:**
 ```
-Using the loaded voice guide, write a review of [product].
-
-Capture the voice by:
-- Starting with personal context (why you tested this)
-- Blending technical specs with practical implications
-- Using the transition patterns between paragraphs
-- Including the "human tells" - parenthetical asides, mid-thought corrections
-- Ending with the closing patterns shown in examples
+Load C:\writing\voice-models\my-voice\writing_style_my-voice.md 
+and use it to write [content type] about [topic]
 ```
 
-**Email/communication:**
-```
-Write an email about [subject] using the voice patterns from the guide.
+The guide includes validation checklists. After Claude writes, run:
 
-Focus on:
-- Conversational markers appearing naturally
-- Sentence length variation (some punchy, some complex)
-- The hedging/confidence balance shown in statistics
-- First-person usage matching the corpus frequency
+```
+Check what you just wrote against the style guide validation checklist.
+Report any violations.
 ```
 
-### Validation After Writing
+### Critical Validation Rules
 
-The guide includes statistical targets. After writing, check:
+The guide flags these as must-pass:
+- Zero AI slop words (delve, leverage, unlock, seamless, robust)
+- Zero em-dashes if corpus doesn't use them
+- British/American spelling consistency
+- Equipment named specifically (not "the product")
 
-```
-Review what you just wrote against the voice guide metrics:
-- Does sentence length variation match the target standard deviation?
-- Is first-person frequency within the expected range?
-- Are conversational markers present but not overused?
-- Does the rhythm feel like the extended examples?
-```
+### Voice Match Validation
 
-## What Gets Analysed
+The guide checks these as should-pass:
+- First-person frequency matches target (~0.8 per 100 words typical)
+- Sentence length varies wildly (5-word to 40-word sentences)
+- Honest caveats present ("It's not perfect", "I wish I'd...")
+- Opening follows corpus patterns
 
-### Core Voice Patterns
-- **Vocabulary** - Word choice, British/American markers, domain specificity
-- **Sentence structure** - Length distribution, openers, complexity patterns
-- **Voice markers** - First-person usage, hedging language, conversational markers
-- **Punctuation** - Dash types, comma density, parenthetical frequency
+## Analysis Output
 
-### Argument & Flow Patterns
-- **Argument flow** - How you open, build, and close arguments
-- **Paragraph transitions** - How ideas connect across paragraphs
-- **Conversational devices** - "look", "frankly", "actually" and when they appear
+The tool generates these JSON files in `corpus-name/analysis/`:
 
-### Micro-Rhythm Detection
-The guide annotates examples with invisible patterns that make writing feel human:
-- Mid-thought pivots (comma before "and", "but", "so")
-- Present-tense immediacy ("Right now, it's...")
-- Embedded uncertainty ("I think", "probably")
-- Casual sentence starters ("So,", "And,", "But,")
-- Parenthetical asides
-- Punchy fragments contrasting with longer sentences
+**Core Analysis:**
+- `vocabulary.json` - Word choice, domain terms, British/American markers
+- `sentence.json` - Length distribution, complexity patterns
+- `voice.json` - First-person usage, hedging language, conversational markers
+- `paragraph.json` - Structure and transition patterns
+- `punctuation.json` - Dash types, comma density, parenthetical frequency
 
-## Output Structure
+**v1.4.0 Additions:**
+- `vocabulary-tiers.json` - AI slop detection, formality scoring with alternatives
+- `phrase-library.json` - 50+ extracted phrases organized by type
 
-```
-your-output-directory/
-└── corpus-name/
-    ├── articles/                              # Collected markdown files
-    ├── corpus.json                            # Metadata
-    ├── analysis/                              # JSON analysis files
-    │   ├── vocabulary.json
-    │   ├── sentence.json
-    │   ├── voice.json
-    │   ├── paragraph.json
-    │   ├── punctuation.json
-    │   ├── function-words.json
-    │   ├── argument-flow.json
-    │   └── paragraph-transitions.json
-    └── writing_style_[name]_narrative.md      # The voice guide
-```
+**Advanced Analysis:**
+- `function-words.json` - Z-scores for style fingerprinting
+- `anti-mechanical.json` - Naturalness scoring
+- `argument-flow.json` - How arguments open, build, close
+- `paragraph-transitions.json` - Cross-paragraph connection patterns
+- `specificity-patterns.json` - Possessive vs generic references
 
-## Minimum Corpus Size
+## Minimum Corpus Requirements
 
 - **Minimum:** 15,000 words (~20 articles)
-- **Recommended:** 30,000 words
+- **Recommended:** 30,000 words (~40 articles)
 - **Ideal:** 50,000+ words
 
-Below 15k words, statistical patterns become unreliable.
+Below 15k words, statistical patterns become unreliable. The phrase library needs volume to find frequently-used patterns.
 
-## Tools Reference
+## MCP Tools Reference
 
 ### collect_corpus
 
-| Parameter | Required | Description |
-|-----------|----------|-------------|
-| `sitemap_url` | Yes | XML sitemap URL |
-| `output_name` | Yes | Corpus identifier (e.g., "my-voice") |
-| `output_dir` | Yes | Directory to store corpus |
-| `max_articles` | No | Limit (default: 100) |
-| `article_pattern` | No | Regex filter for URLs |
+Crawls sitemap and collects clean writing corpus.
+
+**Parameters:**
+- `sitemap_url` (required) - XML sitemap URL
+- `output_name` (required) - Corpus identifier
+- `output_dir` (required) - Storage directory
+- `max_articles` (optional) - Limit, default 100
+- `article_pattern` (optional) - Regex URL filter
 
 ### analyze_corpus
 
-| Parameter | Required | Description |
-|-----------|----------|-------------|
-| `corpus_name` | Yes | Name from collect_corpus |
-| `corpus_dir` | Yes | Directory containing corpus |
-| `analysis_type` | No | full, quick, vocabulary, syntax |
+Runs linguistic analysis on collected corpus.
+
+**Parameters:**
+- `corpus_name` (required) - Name from collect_corpus
+- `corpus_dir` (required) - Directory containing corpus
+- `analysis_type` (optional) - full, quick, vocabulary, syntax (default: full)
+
+### generate_style_guide
 
-### generate_narrative_guide
+Generates v4 executable style guide.
 
-| Parameter | Required | Description |
-|-----------|----------|-------------|
-| `corpus_name` | Yes | Name from analyze_corpus |
-| `corpus_dir` | Yes | Directory containing corpus |
+**Parameters:**
+- `corpus_name` (required) - Name from analyze_corpus
+- `corpus_dir` (required) - Directory containing analysis
 
 ## Development
 
@@ -253,13 +226,37 @@ npm install
 npm run build
 ```
 
-## Limitations
+Local development mode in Claude Desktop config:
+
+```json
+{
+  "mcpServers": {
+    "voice-analysis": {
+      "command": "node",
+      "args": ["C:\\path\\to\\mcp-server-voice-analysis\\dist\\index.js"]
+    }
+  }
+}
+```
+
+## Known Limitations
+
+- Needs XML sitemap (RSS feeds not supported)
+- Works best with single-author content
+- Mixed authorship weakens statistical signals
+- Heavily edited content produces less distinct voice patterns
+- Transition phrase detection currently returns sparse results (being improved)
+
+## What's Next
 
-- Requires XML sitemap (RSS feeds not currently supported)
-- Works best with consistent single-author content
-- Mixed authorship or heavily edited content produces weaker signals
-- The approach is experimental - results vary by writing style
+v1.5.0 planned features:
+- Automated text validation against corpus
+- Real-time writing feedback
+- Custom forbidden vocabulary per corpus
+- Improved transition phrase detection
 
 ---
 
-Apache License 2.0 - [Houtini.ai](https://houtini.ai)
+**License:** Apache 2.0  
+**Author:** [Houtini](https://houtini.ai)  
+**Version:** 1.4.0

package/dist/analyzers/phrase-extraction.d.ts

@@ -0,0 +1,30 @@
+/**
+ * Phrase Extraction Analysis
+ *
+ * Extracts frequently-used phrases for direct imitation.
+ * Focuses on opening patterns, transitions, equipment references, and caveats.
+ */
+export interface PhraseExample {
+    phrase: string;
+    count: number;
+    context?: string;
+}
+export interface PhraseLibrary {
+    openingPatterns: {
+        personalStory: PhraseExample[];
+        directAction: PhraseExample[];
+        protectiveWarning: PhraseExample[];
+    };
+    transitionPhrases: PhraseExample[];
+    equipmentReferences: {
+        withPossessive: PhraseExample[];
+        generic: PhraseExample[];
+    };
+    caveatPhrases: PhraseExample[];
+    totalPhrases: number;
+}
+/**
+ * Main phrase extraction function
+ */
+export declare function extractPhrases(text: string): PhraseLibrary;
+//# sourceMappingURL=phrase-extraction.d.ts.map

package/dist/analyzers/phrase-extraction.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"phrase-extraction.d.ts","sourceRoot":"","sources":["../../src/analyzers/phrase-extraction.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AAIH,MAAM,WAAW,aAAa;IAC5B,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,EAAE,MAAM,CAAC;IACd,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,aAAa;IAC5B,eAAe,EAAE;QACf,aAAa,EAAE,aAAa,EAAE,CAAC;QAC/B,YAAY,EAAE,aAAa,EAAE,CAAC;QAC9B,iBAAiB,EAAE,aAAa,EAAE,CAAC;KACpC,CAAC;IACF,iBAAiB,EAAE,aAAa,EAAE,CAAC;IACnC,mBAAmB,EAAE;QACnB,cAAc,EAAE,aAAa,EAAE,CAAC;QAChC,OAAO,EAAE,aAAa,EAAE,CAAC;KAC1B,CAAC;IACF,aAAa,EAAE,aAAa,EAAE,CAAC;IAC/B,YAAY,EAAE,MAAM,CAAC;CACtB;AAkLD;;GAEG;AACH,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,GAAG,aAAa,CAsB1D"}

package/dist/analyzers/phrase-extraction.js

@@ -0,0 +1,175 @@
+/**
+ * Phrase Extraction Analysis
+ *
+ * Extracts frequently-used phrases for direct imitation.
+ * Focuses on opening patterns, transitions, equipment references, and caveats.
+ */
+import nlp from 'compromise';
+/**
+ * Extract n-grams (2-6 word phrases) from text
+ */
+function extractNGrams(text, minLength = 2, maxLength = 6) {
+    const words = text.toLowerCase()
+        .replace(/[—–]/g, '-')
+        .split(/\s+/)
+        .filter(w => w.length > 0);
+    const ngramCounts = new Map();
+    for (let n = minLength; n <= maxLength; n++) {
+        for (let i = 0; i <= words.length - n; i++) {
+            const ngram = words.slice(i, i + n).join(' ');
+            ngramCounts.set(ngram, (ngramCounts.get(ngram) || 0) + 1);
+        }
+    }
+    return ngramCounts;
+}
+/**
+ * Extract opening patterns from paragraphs
+ */
+function extractOpeningPatterns(text) {
+    const paragraphs = text.split(/\n\n+/);
+    const openingSentences = [];
+    for (const para of paragraphs) {
+        const sentences = para.split(/[.!?]+/).filter(s => s.trim().length > 10);
+        if (sentences.length > 0) {
+            openingSentences.push(sentences[0].trim());
+        }
+    }
+    const personalStory = [];
+    const directAction = [];
+    const protectiveWarning = [];
+    const personalMarkers = /\b(i've|i'm|i was|i have|for me|my|when i|how i)\b/i;
+    const actionMarkers = /\b(right|so|now|let's|first|start|here's|okay)\b/i;
+    const warningMarkers = /\b(before|make sure|important|note|remember|warning|careful)\b/i;
+    for (const sentence of openingSentences) {
+        const lower = sentence.toLowerCase();
+        if (personalMarkers.test(lower)) {
+            personalStory.push({ phrase: sentence, count: 1 });
+        }
+        else if (actionMarkers.test(lower)) {
+            directAction.push({ phrase: sentence, count: 1 });
+        }
+        else if (warningMarkers.test(lower)) {
+            protectiveWarning.push({ phrase: sentence, count: 1 });
+        }
+    }
+    return { personalStory, directAction, protectiveWarning };
+}
+/**
+ * Extract transition phrases (between paragraphs or at sentence starts)
+ */ function extractTransitionPhrases(text) {
+    const sentences = text.split(/[.!?]+/).filter(s => s.trim().length > 10);
+    const transitionMarkers = [
+        'at this', 'once you', 'once we', 'at that', 'after that', 'from there',
+        'next up', 'moving on', 'the key', 'the thing', 'in fact', 'actually',
+        'however', 'though', 'still', 'meanwhile', 'alternatively',
+        'for instance', 'for example', 'essentially', 'basically'
+    ];
+    const transitions = new Map();
+    for (const sentence of sentences) {
+        const lower = sentence.toLowerCase().trim();
+        for (const marker of transitionMarkers) {
+            if (lower.startsWith(marker)) {
+                const match = sentence.match(new RegExp(`^${marker}[^,]*`, 'i'));
+                if (match) {
+                    const phrase = match[0].trim();
+                    transitions.set(phrase, (transitions.get(phrase) || 0) + 1);
+                }
+            }
+        }
+    }
+    return Array.from(transitions.entries())
+        .filter(([_, count]) => count >= 1)
+        .map(([phrase, count]) => ({ phrase, count }))
+        .sort((a, b) => b.count - a.count);
+}
+/**
+ * Extract equipment references with possessives vs generic
+ */
+function extractEquipmentReferences(text) {
+    const equipmentWords = [
+        'pc', 'rig', 'card', 'gpu', 'block', 'pad', 'thermal', 'cooler',
+        'radiator', 'pump', 'fan', 'case', 'system', 'setup', 'hardware',
+        'device', 'product', 'equipment', 'unit', 'tool'
+    ];
+    const withPossessive = new Map();
+    const generic = new Map();
+    const doc = nlp(text);
+    for (const word of equipmentWords) {
+        const possessiveMatches = doc.match(`(my|our) ${word}`);
+        possessiveMatches.forEach((match) => {
+            const phrase = match.text().toLowerCase();
+            withPossessive.set(phrase, (withPossessive.get(phrase) || 0) + 1);
+        });
+        const genericMatches = doc.match(`(the|this|that|a) ${word}`);
+        genericMatches.forEach((match) => {
+            const phrase = match.text().toLowerCase();
+            generic.set(phrase, (generic.get(phrase) || 0) + 1);
+        });
+    }
+    return {
+        withPossessive: Array.from(withPossessive.entries())
+            .map(([phrase, count]) => ({ phrase, count }))
+            .sort((a, b) => b.count - a.count)
+            .slice(0, 30),
+        generic: Array.from(generic.entries())
+            .map(([phrase, count]) => ({ phrase, count }))
+            .sort((a, b) => b.count - a.count)
+            .slice(0, 20)
+    };
+}
+/**
+ * Extract caveat and honesty phrases
+ */
+function extractCaveatPhrases(text) {
+    const caveatMarkers = [
+        "it isn't", "it's not", "not perfect", "not ideal", "could be better",
+        "i wish", "would have", "should have", "probably", "might not",
+        "may not", "doesn't always", "won't always", "can be", "tends to",
+        "in my case", "for me", "your mileage"
+    ];
+    const caveats = new Map();
+    const sentences = text.split(/[.!?]+/).filter(s => s.trim().length > 10);
+    for (const sentence of sentences) {
+        const lower = sentence.toLowerCase();
+        for (const marker of caveatMarkers) {
+            if (lower.includes(marker)) {
+                const match = sentence.match(new RegExp(`[^.!?]*${marker}[^.!?]*`, 'i'));
+                if (match) {
+                    const phrase = match[0].trim();
+                    if (phrase.length > 10 && phrase.length < 100) {
+                        caveats.set(phrase, (caveats.get(phrase) || 0) + 1);
+                    }
+                }
+            }
+        }
+    }
+    return Array.from(caveats.entries())
+        .filter(([_, count]) => count >= 1)
+        .map(([phrase, count]) => ({ phrase, count }))
+        .sort((a, b) => b.count - a.count)
+        .slice(0, 20);
+}
+/**
+ * Main phrase extraction function
+ */
+export function extractPhrases(text) {
+    const openingPatterns = extractOpeningPatterns(text);
+    const transitionPhrases = extractTransitionPhrases(text);
+    const equipmentReferences = extractEquipmentReferences(text);
+    const caveatPhrases = extractCaveatPhrases(text);
+    const totalPhrases = openingPatterns.personalStory.length +
+        openingPatterns.directAction.length +
+        openingPatterns.protectiveWarning.length +
+        transitionPhrases.length +
+        equipmentReferences.withPossessive.length +
+        equipmentReferences.generic.length +
+        caveatPhrases.length;
+    return {
+        openingPatterns,
+        transitionPhrases,
+        equipmentReferences,
+        caveatPhrases,
+        totalPhrases
+    };
+}
+//# sourceMappingURL=phrase-extraction.js.map

package/dist/analyzers/phrase-extraction.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"phrase-extraction.js","sourceRoot":"","sources":["../../src/analyzers/phrase-extraction.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AAEH,OAAO,GAAG,MAAM,YAAY,CAAC;AAuB7B;;GAEG;AACH,SAAS,aAAa,CAAC,IAAY,EAAE,YAAoB,CAAC,EAAE,YAAoB,CAAC;IAC/E,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,EAAE;SAC7B,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,KAAK,CAAC,KAAK,CAAC;SACZ,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAE7B,MAAM,WAAW,GAAG,IAAI,GAAG,EAAkB,CAAC;IAE9C,KAAK,IAAI,CAAC,GAAG,SAAS,EAAE,CAAC,IAAI,SAAS,EAAE,CAAC,EAAE,EAAE,CAAC;QAC5C,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC;YAC3C,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC9C,WAAW,CAAC,GAAG,CAAC,KAAK,EAAE,CAAC,WAAW,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QAC5D,CAAC;IACH,CAAC;IAED,OAAO,WAAW,CAAC;AACrB,CAAC;AAED;;GAEG;AACH,SAAS,sBAAsB,CAAC,IAAY;IAK1C,MAAM,UAAU,GAAG,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IACvC,MAAM,gBAAgB,GAAa,EAAE,CAAC;IAEtC,KAAK,MAAM,IAAI,IAAI,UAAU,EAAE,CAAC;QAC9B,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC;QACzE,IAAI,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACzB,gBAAgB,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;QAC7C,CAAC;IACH,CAAC;IAED,MAAM,aAAa,GAAoB,EAAE,CAAC;IAC1C,MAAM,YAAY,GAAoB,EAAE,CAAC;IACzC,MAAM,iBAAiB,GAAoB,EAAE,CAAC;IAE9C,MAAM,eAAe,GAAG,qDAAqD,CAAC;IAC9E,MAAM,aAAa,GAAG,mDAAmD,CAAC;IAC1E,MAAM,cAAc,GAAG,iEAAiE,CAAC;IAEzF,KAAK,MAAM,QAAQ,IAAI,gBAAgB,EAAE,CAAC;QACxC,MAAM,KAAK,GAAG,QAAQ,CAAC,WAAW,EAAE,CAAC;QAErC,IAAI,eAAe,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YAChC,aAAa,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,QAAQ,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAC;QACrD,CAAC;aAAM,IAAI,aAAa,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YACrC,YAAY,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,QAAQ,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAC;QACpD,CAAC;aAAM,IAAI,cAAc,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YACtC,iBAAiB,CAAC,IAAI,CAAC,EAAE,MAAM,EAAE,QAAQ,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAC;QACzD,CAAC;IACH,CAAC;IAED,OAAO,EAAE,aAAa,EAAE,YAAY,EAAE,iBAAiB,EAAE,CAAC;AAC5D,CAAC;AAED;;GAEG,CAAA,SAAS,wBAAwB,CAAC,IAAY;IAC/C,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC;IACzE,MAAM,iBAAiB,GAAG;QACxB,SAAS,EAAE,UAAU,EAAE,SAAS,EAAE,SAAS,EAAE,YAAY,EAAE,YAAY;QACvE,SAAS,EAAE,WAAW,EAAE,SAAS,EAAE,WAAW,EAAE,SAAS,EAAE,UAAU;QACrE,SAAS,EAAE,QAAQ,EAAE,OAAO,EAAE,WAAW,EAAE,eAAe;QAC1D,cAAc,EAAE,aAAa,EAAE,aAAa,EAAE,WAAW;KAC1D,CAAC;IAEF,MAAM,WAAW,GAAG,IAAI,GAAG,EAAkB,CAAC;IAE9C,KAAK,MAAM,QAAQ,IAAI,SAAS,EAAE,CAAC;QACjC,MAAM,KAAK,GAAG,QAAQ,CAAC,WAAW,EAAE,CAAC,IAAI,EAAE,CAAC;QAE5C,KAAK,MAAM,MAAM,IAAI,iBAAiB,EAAE,CAAC;YACvC,IAAI,KAAK,CAAC,UAAU,CAAC,MAAM,CAAC,EAAE,CAAC;gBAC7B,MAAM,KAAK,GAAG,QAAQ,CAAC,KAAK,CAAC,IAAI,MAAM,CAAC,IAAI,MAAM,OAAO,EAAE,GAAG,CAAC,CAAC,CAAC;gBACjE,IAAI,KAAK,EAAE,CAAC;oBACV,MAAM,MAAM,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;oBAC/B,WAAW,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC,WAAW,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;gBAC9D,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,WAAW,CAAC,OAAO,EAAE,CAAC;SACrC,MAAM,CAAC,CAAC,CAAC,CAAC,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,KAAK,IAAI,CAAC,CAAC;SAClC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,CAAC;SAC7C,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC;AACvC,CAAC;AAED;;GAEG;AACH,SAAS,0BAA0B,CAAC,IAAY;IAI9C,MAAM,cAAc,GAAG;QACrB,IAAI,EAAE,KAAK,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,KAAK,EAAE,SAAS,EAAE,QAAQ;QAC/D,UAAU,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,QAAQ,EAAE,OAAO,EAAE,UAAU;QAChE,QAAQ,EAAE,SAAS,EAAE,WAAW,EAAE,MAAM,EAAE,MAAM;KACjD,CAAC;IAEF,MAAM,cAAc,GAAG,IAAI,GAAG,EAAkB,CAAC;IACjD,MAAM,OAAO,GAAG,IAAI,GAAG,EAAkB,CAAC;IAE1C,MAAM,GAAG,GAAG,GAAG,CAAC,IAAI,CAAC,CAAC;IAEtB,KAAK,MAAM,IAAI,IAAI,cAAc,EAAE,CAAC;QAClC,MAAM,iBAAiB,GAAG,GAAG,CAAC,KAAK,CAAC,YAAY,IAAI,EAAE,CAAC,CAAC;QACxD,iBAAiB,CAAC,OAAO,CAAC,CAAC,KAAU,EAAE,EAAE;YACvC,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;YAC1C,cAAc,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC,cAAc,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QACpE,CAAC,CAAC,CAAC;QAEH,MAAM,cAAc,GAAG,GAAG,CAAC,KAAK,CAAC,qBAAqB,IAAI,EAAE,CAAC,CAAC;QAC9D,cAAc,CAAC,OAAO,CAAC,CAAC,KAAU,EAAE,EAAE;YACpC,MAAM,MAAM,GAAG,KAAK,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;YAC1C,OAAO,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;QACtD,CAAC,CAAC,CAAC;IACL,CAAC;IAED,OAAO;QACL,cAAc,EAAE,KAAK,CAAC,IAAI,CAAC,cAAc,CAAC,OAAO,EAAE,CAAC;aACjD,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,CAAC;aAC7C,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC;aACjC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC;QACf,OAAO,EAAE,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC;aACnC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,CAAC;aAC7C,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC;aACjC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC;KAChB,CAAC;AACJ,CAAC;AAED;;GAEG;AACH,SAAS,oBAAoB,CAAC,IAAY;IACxC,MAAM,aAAa,GAAG;QACpB,UAAU,EAAE,UAAU,EAAE,aAAa,EAAE,WAAW,EAAE,iBAAiB;QACrE,QAAQ,EAAE,YAAY,EAAE,aAAa,EAAE,UAAU,EAAE,WAAW;QAC9D,SAAS,EAAE,gBAAgB,EAAE,cAAc,EAAE,QAAQ,EAAE,UAAU;QACjE,YAAY,EAAE,QAAQ,EAAE,cAAc;KACvC,CAAC;IAEF,MAAM,OAAO,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC1C,MAAM,SAAS,GAAG,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,EAAE,CAAC,CAAC;IAEzE,KAAK,MAAM,QAAQ,IAAI,SAAS,EAAE,CAAC;QACjC,MAAM,KAAK,GAAG,QAAQ,CAAC,WAAW,EAAE,CAAC;QAErC,KAAK,MAAM,MAAM,IAAI,aAAa,EAAE,CAAC;YACnC,IAAI,KAAK,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC;gBAC3B,MAAM,KAAK,GAAG,QAAQ,CAAC,KAAK,CAAC,IAAI,MAAM,CAAC,UAAU,MAAM,SAAS,EAAE,GAAG,CAAC,CAAC,CAAC;gBACzE,IAAI,KAAK,EAAE,CAAC;oBACV,MAAM,MAAM,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;oBAC/B,IAAI,MAAM,CAAC,MAAM,GAAG,EAAE,IAAI,MAAM,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;wBAC9C,OAAO,CAAC,GAAG,CAAC,MAAM,EAAE,CAAC,OAAO,CAAC,GAAG,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;oBACtD,CAAC;gBACH,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC;SACjC,MAAM,CAAC,CAAC,CAAC,CAAC,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,KAAK,IAAI,CAAC,CAAC;SAClC,GAAG,CAAC,CAAC,CAAC,MAAM,EAAE,KAAK,CAAC,EAAE,EAAE,CAAC,CAAC,EAAE,MAAM,EAAE,KAAK,EAAE,CAAC,CAAC;SAC7C,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC;SACjC,KAAK,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;AAClB,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,cAAc,CAAC,IAAY;IACzC,MAAM,eAAe,GAAG,sBAAsB,CAAC,IAAI,CAAC,CAAC;IACrD,MAAM,iBAAiB,GAAG,wBAAwB,CAAC,IAAI,CAAC,CAAC;IACzD,MAAM,mBAAmB,GAAG,0BAA0B,CAAC,IAAI,CAAC,CAAC;IAC7D,MAAM,aAAa,GAAG,oBAAoB,CAAC,IAAI,CAAC,CAAC;IAEjD,MAAM,YAAY,GAChB,eAAe,CAAC,aAAa,CAAC,MAAM;QACpC,eAAe,CAAC,YAAY,CAAC,MAAM;QACnC,eAAe,CAAC,iBAAiB,CAAC,MAAM;QACxC,iBAAiB,CAAC,MAAM;QACxB,mBAAmB,CAAC,cAAc,CAAC,MAAM;QACzC,mBAAmB,CAAC,OAAO,CAAC,MAAM;QAClC,aAAa,CAAC,MAAM,CAAC;IAEvB,OAAO;QACL,eAAe;QACf,iBAAiB;QACjB,mBAAmB;QACnB,aAAa;QACb,YAAY;KACb,CAAC;AACJ,CAAC"}

package/dist/analyzers/vocabulary-tiers.d.ts

@@ -0,0 +1,27 @@
+/**
+ * Vocabulary Tier Analysis
+ *
+ * Identifies formal vs casual vocabulary to help writers avoid
+ * using words that are too formal for their authentic voice.
+ */
+export interface VocabularyTier {
+    word: string;
+    count: number;
+    category: 'verb' | 'adjective' | 'adverb';
+    formality: 'formal' | 'ai-slop';
+    suggestedAlternatives: string[];
+}
+export interface VocabularyTierAnalysis {
+    formalVerbs: VocabularyTier[];
+    formalAdjectives: VocabularyTier[];
+    formalAdverbs: VocabularyTier[];
+    aiSlop: VocabularyTier[];
+    totalFormalWords: number;
+    formalityScore: number;
+    recommendations: string[];
+}
+/**
+ * Analyze vocabulary formality in text
+ */
+export declare function analyzeVocabularyTiers(text: string): VocabularyTierAnalysis;
+//# sourceMappingURL=vocabulary-tiers.d.ts.map