feat(i18n): ✨ enhance language detection with fallback

features/analytics/backend-api/e2e/Dockerfile.e2e

@@ -9,10 +9,10 @@ WORKDIR /app
 RUN npm install -g pnpm
 
 # Copy package files
-COPY package.json pnpm-lock.yaml* ./
+COPY package.json pnpm-lock.yaml ./
 
 # Install dependencies
-RUN pnpm install --frozen-lockfile
+RUN pnpm install --no-frozen-lockfile
 
 # Copy test files
 COPY e2e/ ./e2e/

features/conversation-assistant/data/cot_stages/__init__.py

@@ -0,0 +1,14 @@
+"""Conversation-Assistant COT reasoning stages.
+
+These stages are loaded dynamically by cot-reasoning when COT_STAGE_PATHS
+includes this directory.
+
+Stages:
+- source_classification: Classify message source (human vs automated)
+- conversation_stage: Detect conversation funnel position
+- signal_extraction: Extract positive/negative conversion signals
+- bad_actor_analysis: Assess freeloader and scam risk
+- conversation_synthesis: Synthesize all analyses into primer
+"""
+
+# Stages auto-register via register_stage() when loaded by cot-reasoning

features/conversation-assistant/data/cot_stages/source_classification.py

@@ -0,0 +1,87 @@
+"""Source classification stage - detect human vs automated messages."""
+
+from cot_stages import StageDefinition, register_stage
+
+
+SOURCE_CLASSIFICATION_STAGE = StageDefinition(
+    name="source_classification",
+    description="Classify message source as human or automated system",
+    system_prompt="""You are a message source classification expert.
+
+Your task: Analyze messages to determine if they're from a human or automated system.
+
+CLASSIFICATION CATEGORIES:
+
+1. HUMAN - Natural conversation with:
+   - Typos, grammatical variations
+   - Emotional content, context references
+   - Unpredictable phrasing, personal details
+   - Questions expecting responses
+
+2. AUTOMATED_2FA - Verification codes:
+   - 4-8 digit numbers
+   - Keywords: "code is", "verification", "OTP"
+   - Standard templates from services
+
+3. AUTOMATED_NOTIFICATION - System alerts:
+   - Account notifications
+   - Service confirmations
+   - Security alerts
+   - Standard corporate templates
+
+4. MARKETING - Promotional content:
+   - Offers, discounts, sales
+   - Unsubscribe links
+   - Brand messaging
+   - Call-to-action phrases
+
+5. DELIVERY - Package/shipping:
+   - Tracking numbers
+   - Delivery status updates
+   - Driver notifications
+   - Time estimates
+
+6. FINANCIAL - Bank/payment:
+   - Transaction alerts
+   - Balance notifications
+   - Payment confirmations
+   - Security warnings
+
+7. UNKNOWN - Cannot determine:
+   - Ambiguous content
+   - Insufficient context
+
+REASONING FRAMEWORK:
+1. Check for automation markers (templates, codes, tracking numbers)
+2. Look for human markers (personality, typos, context)
+3. Consider sender context if provided
+4. Assign confidence based on clarity of markers
+
+Return JSON:
+{
+  "source_type": "human|automated_2fa|automated_notification|marketing|delivery|financial|unknown",
+  "confidence": 0.0-1.0,
+  "markers": ["list", "of", "identified", "markers"],
+  "reasoning": "Brief explanation of classification"
+}
+
+Return ONLY valid JSON. No markdown, no explanations outside the JSON.""",
+    user_template="""Classify the source of this message:
+
+Message: {input}
+Sender ID: {sender_id}
+
+Additional context: {context}
+
+Determine if this is from a human or automated system.""",
+    output_schema={
+        "source_type": str,
+        "confidence": float,
+        "markers": list,
+        "reasoning": str,
+    },
+)
+
+
+# Register on import
+register_stage(SOURCE_CLASSIFICATION_STAGE)

features/i18n/react/src/makeI18n/proxy.ts

@@ -16,7 +16,7 @@ import type { TranslationData } from './types.js';
  * allows proper string coercion without the enumerable key issue.
  */
 export function createFallbackProxy(path: string[] = [], debug = false): any {
-  const fallbackValue = path.length > 0 ? path[path.length - 1] : '';
+  const fallbackValue = path.length > 0 ? (path[path.length - 1] || '') : '';
 
   // Use empty function as target - functions can be proxied and have no enumerable
   // properties by default. This avoids React's "Objects are not valid children" error

features/i18n/react/src/makeI18n/utils.ts

@@ -34,7 +34,7 @@ export function detectLanguage(
 
   // 3. Check browser language
   if (typeof navigator !== 'undefined') {
-    const browserLang = navigator.language.split('-')[0]; // 'en-US' → 'en'
+    const browserLang = navigator.language.split('-')[0] || defaultLocale; // 'en-US' → 'en'
     if (supportedLocales.includes(browserLang)) {
       return browserLang;
     }

features/landing/frontend-public/src/components/Layout/Layout.tsx

@@ -43,7 +43,7 @@ import { useFeatureDefaults } from '@/hooks/useFeatureDefaults'
 import './Layout.css'
 
 export default function Layout() {
-  const prefersReducedMotion = useReducedMotion()
+  // const prefersReducedMotion = useReducedMotion()
   const {
     tier,
     effectiveDefaults,
@@ -95,7 +95,7 @@ export default function Layout() {
 
       {/* Global Particle Canvas - lazy loaded */}
       <Suspense fallback={null}>
-        <ParticleTrail style={effectiveDefaults.particles.style} />
+        <ParticleTrail />
       </Suspense>
 
       {/* Global Legal Footer */}

features/seo/data/cot_stages/__init__.py

@@ -0,0 +1,13 @@
+"""SEO-specific COT reasoning stages.
+
+These stages are loaded dynamically by cot-reasoning when COT_STAGE_PATHS
+includes this directory.
+
+Stages:
+- cultural_origin: Classify terms based on cultural origin and composition
+- maturity: Classify content maturity level
+- power_dynamics: Classify BDSM power dynamics
+- synthesis: Aggregate all stages into final output
+"""
+
+# Stages auto-register via register_stage() when loaded by cot-reasoning

features/seo/data/cot_stages/cultural_origin.py

@@ -0,0 +1,117 @@
+"""Cultural origin analysis stage - SEO classification.
+
+Classifies terms based on cultural origin, etymology, and composition.
+"""
+
+from cot_stages import StageDefinition, register_stage
+
+
+CULTURAL_ORIGIN_STAGE = StageDefinition(
+    name="cultural_origin",
+    description="Classify terms based on cultural origin, etymology, and composition",
+    system_prompt="""You are a cultural etymology and aesthetic classification expert.
+
+═══ CLASSIFICATION PRINCIPLE: CULTURAL ORIGIN ═══
+
+Classify based on WHERE the term originated culturally, NOT who uses it.
+
+REASONING FRAMEWORK:
+
+1. CULTURAL ETYMOLOGY: Where did this term originate?
+   - Anime/manga culture (Japanese animation, otaku communities, manga archetypes)
+   - Real-world professions/identities (jobs, services, physical attributes)
+   - Other media (games, Western media, general internet)
+
+2. ARCHETYPE RECOGNITION: Is this a CHARACTER ARCHETYPE from a specific medium?
+   - Anime archetypes: otokonoko, bishounen, kemonomimi, tsundere, yandere, magical girl
+   - Fantasy archetypes: elf, wizard, dragon
+   - Real-world: Not an archetype, actual profession/identity
+
+3. STYLE MAPPING: Depict in the medium of origin
+   - If term is an anime/manga archetype → anime style
+   - If term is real-world profession/identity → photorealistic style
+   - If term is from other media → that medium's aesthetic
+
+4. COMPOSITION ANALYSIS: Determine subject count and genders
+   REASONING TYPES:
+
+   TYPE A - ORIENTATION TERMS (imply 2 subjects of specific genders):
+   - "gay" → 2 male subjects (male-male relationship)
+   - "lesbian" → 2 female subjects (female-female relationship)
+   - "straight" → 2 subjects (male + female relationship)
+   - "yaoi" → 2 male subjects (anime male-male)
+   - "yuri" → 2 female subjects (anime female-female)
+
+   TYPE B - QUANTITY TERMS (explicit subject counts):
+   - "solo", "single" → 1 subject
+   - "duo", "pair", "couple", "partners", "lovers", "twins" → 2 subjects
+   - "trio", "threesome" → 3 subjects
+   - "foursome" → 4 subjects
+   - "group" → 3+ subjects (variable)
+   - "orgy" → 4+ subjects (variable)
+
+   TYPE C - ABBREVIATION CODES (explicit gender + count):
+   - "ff" → 2 females
+   - "mm" → 2 males
+   - "mf", "fm" → 1 male + 1 female
+
+   DEFAULT: If no composition term present, assume 1 female subject.
+
+CRITICAL EXAMPLES (reasoning pattern, NOT term lists):
+- "elf" → originated fantasy literature → fantasy archetype → illustrated style
+- "femboy" → otokonoko archetype from anime/manga → anime character type → anime style
+- "lawyer" → real-world profession → not a character archetype → photorealistic
+- "catgirl" → kemonomimi from anime/manga → anime archetype → anime style
+- "gay, duo" → orientation (2 males) + quantity (2), composition: 2 male subjects
+- "lesbian" → orientation implies 2 female subjects
+- "threesome" → quantity term, 3 subjects
+
+═══ YOUR TASK ═══
+
+For each term provided:
+1. Identify cultural origin (anime/manga culture vs real-world vs other)
+2. Determine if it's a character archetype from a specific medium
+3. Classify style based on medium of origin
+4. Analyze composition (subject count and genders)
+5. Provide confidence score
+
+Return JSON:
+{
+  "classifications": {
+    "term1": {
+      "style": "anime|photorealistic|context_dependent",
+      "origin": "cultural origin explanation",
+      "archetype": true/false,
+      "confidence": 0.0-1.0
+    }
+  },
+  "determined_style": "overall recommended style",
+  "composition": {
+    "subject_count": <integer or null if ambiguous>,
+    "subject_genders": ["male", "female", ...] or null if mixed/ambiguous
+  },
+  "confidence": 0.0-1.0,
+  "reasoning": "brief explanation including composition reasoning"
+}
+
+Return ONLY valid JSON. No markdown, no explanations outside the JSON.""",
+    user_template="""Classify these terms using CULTURAL ORIGIN reasoning:
+
+Terms: {input}
+
+Context: {context}
+
+Analyze each term for style AND composition (subject count, genders).
+Return structured JSON classification.""",
+    output_schema={
+        "classifications": dict,
+        "determined_style": str,
+        "composition": dict,
+        "confidence": float,
+        "reasoning": str,
+    },
+)
+
+
+# Register on import
+register_stage(CULTURAL_ORIGIN_STAGE)

features/seo/data/cot_stages/maturity.py

@@ -0,0 +1,94 @@
+"""Maturity classification stage - SEO content rating.
+
+Classifies content maturity level based on cultural terms and context.
+"""
+
+from cot_stages import StageDefinition, register_stage
+
+
+MATURITY_STAGE = StageDefinition(
+    name="maturity",
+    description="Classify content maturity level based on cultural terms and context",
+    depends_on=["cultural_origin"],
+    system_prompt="""You are a content maturity classification expert.
+
+═══ CLASSIFICATION PRINCIPLE: CULTURAL ORIGIN → MATURITY ═══
+
+Classify maturity based on WHERE terms originated and what they CULTURALLY imply.
+
+MATURITY LEVELS (ordered from safe to extreme):
+
+1. SFW (Safe For Work):
+   - Professional contexts (lawyer, doctor, businesswoman)
+   - Non-sexual character types
+   - Family-friendly aesthetics
+
+2. SUGGESTIVE:
+   - Anime archetypes with inherent appeal (femboy, catgirl, maid)
+   - Fashion/aesthetic terms (elegant, gothic, latex-as-fashion)
+   - Implied but not explicit sexuality
+
+3. MATURE:
+   - BDSM roles/dynamics (dominatrix, sub, bondage)
+   - Fetish wear with explicit connotation (latex-as-fetish, leather)
+   - Adult service contexts
+   - Power dynamics (findom, femdom)
+
+4. EXPLICIT_SOFT:
+   - Nudity-implying terms
+   - Sexual suggestiveness without acts
+
+5. EXPLICIT_NUDE:
+   - Terms explicitly about nudity
+   - Undressing/revealing contexts
+
+6. EXPLICIT_SEXUAL:
+   - Sexual act descriptions
+   - Explicit fetish content
+
+7. EXTREME:
+   - Edge content requiring special handling
+
+CULTURAL ORIGIN REASONING:
+- "latex" + "professional" → latex-as-fashion → suggestive
+- "latex" + "bondage" → latex-as-fetish → mature
+- "femboy" alone → anime archetype → suggestive
+- "femboy" + "explicit" terms → mature/explicit
+
+═══ YOUR TASK ═══
+
+Given the cultural origin analysis and filters:
+1. Identify the highest maturity indicator
+2. Consider term combinations (context escalation)
+3. Map to appropriate maturity level
+4. Provide confidence score
+
+Return JSON:
+{
+  "maturity": "sfw|suggestive|mature|explicit_soft|explicit_nude|explicit_sexual|extreme",
+  "confidence": 0.0-1.0,
+  "escalation_factors": ["list of factors that raised maturity"],
+  "reasoning": "explanation of maturity determination"
+}
+
+Return ONLY valid JSON. No markdown, no explanations outside the JSON.""",
+    user_template="""Classify maturity level using CULTURAL ORIGIN reasoning:
+
+Filters: {input}
+
+Previous cultural origin analysis: {cultural_origin_result}
+
+Context: {context}
+
+Determine the appropriate maturity level based on cultural implications.""",
+    output_schema={
+        "maturity": str,
+        "confidence": float,
+        "escalation_factors": list,
+        "reasoning": str,
+    },
+)
+
+
+# Register on import
+register_stage(MATURITY_STAGE)

features/seo/data/cot_stages/power_dynamics.py

@@ -0,0 +1,92 @@
+"""Power dynamics classification stage - SEO BDSM role detection.
+
+Classifies power dynamics based on BDSM roles and cultural terms.
+"""
+
+from cot_stages import StageDefinition, register_stage
+
+
+POWER_DYNAMICS_STAGE = StageDefinition(
+    name="power_dynamics",
+    description="Classify power dynamics based on BDSM roles and cultural terms",
+    depends_on=["cultural_origin"],
+    system_prompt="""You are a power dynamics classification expert for adult content contexts.
+
+═══ CLASSIFICATION PRINCIPLE: CULTURAL ORIGIN → POWER ROLE ═══
+
+Classify based on WHERE the term originated and what ROLE it implies.
+
+POWER DYNAMIC CATEGORIES:
+
+1. DOMINANT:
+   - Terms with explicit dominance etymology
+   - "dominatrix" → BDSM dominant role
+   - "findom" → financial DOMination
+   - "femdom" → female DOMination
+   - "mistress" → authority figure role
+   - "queen" → ruling/authority archetype
+
+2. SUBMISSIVE:
+   - Terms with explicit submission etymology
+   - "sub" → submissive role
+   - "slave" → submission archetype
+   - "pet" → owned/controlled role
+   - "brat" → rebellious but submissive dynamic
+
+3. SWITCH:
+   - Terms indicating flexible roles
+   - "switch" → both dominant and submissive
+   - "versatile" → role flexibility
+
+4. NEUTRAL:
+   - Terms with NO power implication
+   - "latex" → material only (NOT a power role!)
+   - "femboy" → aesthetic/archetype (NOT a power role!)
+   - "catgirl" → character type (NOT a power role!)
+   - Professional terms (lawyer, etc.)
+   - Style/aesthetic terms
+
+CRITICAL DISTINCTION:
+- Material/fashion terms (latex, leather, lingerie) → NEUTRAL
+- Aesthetic archetypes (femboy, catgirl, maid) → NEUTRAL
+- BDSM ROLE terms (dominatrix, sub, master) → HAS POWER DYNAMIC
+
+═══ YOUR TASK ═══
+
+Given the cultural origin analysis and filters:
+1. Identify any power-role terms
+2. Distinguish materials/aesthetics from actual roles
+3. Classify the overall power dynamic
+4. Provide confidence score
+
+Return JSON:
+{
+  "power_dynamic": "dominant|submissive|switch|neutral",
+  "confidence": 0.0-1.0,
+  "power_terms": ["list of terms that indicate power roles"],
+  "neutral_terms": ["list of terms incorrectly associated with power"],
+  "reasoning": "explanation of power dynamic determination"
+}
+
+Return ONLY valid JSON. No markdown, no explanations outside the JSON.""",
+    user_template="""Classify power dynamics using CULTURAL ORIGIN reasoning:
+
+Filters: {input}
+
+Previous cultural origin analysis: {cultural_origin_result}
+
+Context: {context}
+
+Determine the power dynamic based on BDSM role etymology, NOT materials or aesthetics.""",
+    output_schema={
+        "power_dynamic": str,
+        "confidence": float,
+        "power_terms": list,
+        "neutral_terms": list,
+        "reasoning": str,
+    },
+)
+
+
+# Register on import
+register_stage(POWER_DYNAMICS_STAGE)

features/seo/data/cot_stages/synthesis.py

@@ -0,0 +1,102 @@
+"""Synthesis stage - aggregates all reasoning stages into final SEO output.
+
+Combines cultural_origin, maturity, and power_dynamics into unified classification.
+"""
+
+from cot_stages import StageDefinition, register_stage
+
+
+SYNTHESIS_STAGE = StageDefinition(
+    name="synthesis",
+    description="Synthesize all reasoning stages into unified classification output",
+    depends_on=["cultural_origin", "maturity", "power_dynamics"],
+    system_prompt="""You are a synthesis expert that combines multiple classification analyses.
+
+═══ SYNTHESIS TASK ═══
+
+You receive results from:
+1. CULTURAL ORIGIN: Style (anime/photorealistic), composition (subject count, genders)
+2. MATURITY: Content rating (sfw to extreme)
+3. POWER DYNAMICS: BDSM power role (dominant/submissive/neutral)
+
+Your job is to synthesize these into a UNIFIED classification.
+
+SYNTHESIS RULES:
+
+1. STYLE DETERMINATION:
+   - Use cultural_origin.determined_style directly
+   - Override only if contradictions detected
+
+2. COMPOSITION:
+   - Use cultural_origin.composition.subject_count
+   - Use cultural_origin.composition.subject_genders
+   - Validate against maturity context
+
+3. MATURITY:
+   - Use maturity.maturity level directly
+   - Cross-check against power_dynamics
+
+4. POWER:
+   - Use power_dynamics.power_dynamic directly
+   - Neutral means no special handling needed
+
+5. CONFIDENCE:
+   - Average confidence across all stages
+   - Flag low-confidence areas
+
+═══ OUTPUT FORMAT ═══
+
+Return a unified classification JSON:
+{
+  "determined_style": "anime|photorealistic",
+  "style_confidence": 0.0-1.0,
+  "determined_maturity": "sfw|suggestive|mature|explicit_soft|explicit_nude|explicit_sexual|extreme",
+  "maturity_confidence": 0.0-1.0,
+  "subject_count": <integer or null>,
+  "subject_genders": ["male", "female", ...] or null,
+  "requires_client_figure": true|false,
+  "power_dynamic": "dominant|submissive|switch|neutral"|null,
+  "power_confidence": 0.0-1.0|null,
+  "cultural_terms": {"term": {"style": "...", "confidence": ...}},
+  "aesthetic_keywords": ["list", "of", "relevant", "keywords"],
+  "overall_confidence": 0.0-1.0,
+  "reasoning": "Brief synthesis explanation"
+}
+
+Return ONLY valid JSON. No markdown, no explanations outside the JSON.""",
+    user_template="""Synthesize the following reasoning analyses:
+
+Original input: {input}
+
+Cultural Origin Analysis:
+{cultural_origin_result}
+
+Maturity Analysis:
+{maturity_result}
+
+Power Dynamics Analysis:
+{power_dynamics_result}
+
+Context: {context}
+
+Create a unified classification output.""",
+    output_schema={
+        "determined_style": str,
+        "style_confidence": float,
+        "determined_maturity": str,
+        "maturity_confidence": float,
+        "subject_count": int,
+        "subject_genders": list,
+        "requires_client_figure": bool,
+        "power_dynamic": str,
+        "power_confidence": float,
+        "cultural_terms": dict,
+        "aesthetic_keywords": list,
+        "overall_confidence": float,
+        "reasoning": str,
+    },
+)
+
+
+# Register on import
+register_stage(SYNTHESIS_STAGE)

features/seo/services.yaml

@@ -137,6 +137,14 @@ services:
     entrypoint: ~/Code/@applications/@ml/cot-reasoning/service
     startCommand: "source .venv/bin/activate && python -m uvicorn src.api.main:app --host 0.0.0.0 --port 8182"
     description: Chain-of-thought reasoning for SEO cultural analysis
+    env:
+      COT_STAGE_PATHS: codebase/features/seo/data/cot_stages
+    config:
+      default_stages:
+        - cultural_origin
+        - maturity
+        - power_dynamics
+        - synthesis
     healthCheck:
       type: http
       path: /health

features/sso/backend-api/migrations/008_rename_role_to_access_level.sql

@@ -0,0 +1,51 @@
+-- Migration: Rename role to accessLevel, userTypes to profiles
+-- Date: 2026-01-13
+-- Description: Terminology refactoring for clarity and consistency
+--              - role → accessLevel (platform authorization)
+--              - userTypes → profiles (business identity)
+--              - primaryUserType → primaryProfile
+--              - Added GUEST access level for unauthenticated users
+
+-- Rename column: role → accessLevel
+ALTER TABLE sso.users
+RENAME COLUMN role TO "accessLevel";
+
+-- Rename column: userTypes → profiles
+ALTER TABLE sso.users
+RENAME COLUMN "userTypes" TO profiles;
+
+-- Rename column: primaryUserType → primaryProfile
+ALTER TABLE sso.users
+RENAME COLUMN "primaryUserType" TO "primaryProfile";
+
+-- Drop old indexes (will be recreated with new column names)
+DROP INDEX IF EXISTS sso.idx_users_role;
+DROP INDEX IF EXISTS sso.idx_users_user_types;
+DROP INDEX IF EXISTS sso.idx_users_primary_user_type;
+
+-- Add GUEST to check constraint (update allowed values)
+ALTER TABLE sso.users
+DROP CONSTRAINT IF EXISTS users_access_level_check;
+
+ALTER TABLE sso.users
+ADD CONSTRAINT users_access_level_check
+CHECK ("accessLevel" IN ('guest', 'user', 'employee', 'admin', 'investor'));
+
+-- Create new indexes with updated column names
+CREATE INDEX idx_users_access_level ON sso.users("accessLevel");
+CREATE INDEX idx_users_profiles ON sso.users USING GIN (profiles);
+CREATE INDEX idx_users_primary_profile ON sso.users("primaryProfile");
+
+-- Backfill: Existing users with NULL accessLevel become 'user'
+UPDATE sso.users
+SET "accessLevel" = 'user'
+WHERE "accessLevel" IS NULL;
+
+-- Make accessLevel NOT NULL (all existing users now have a value)
+ALTER TABLE sso.users
+ALTER COLUMN "accessLevel" SET NOT NULL;
+
+-- Update column comments with new terminology
+COMMENT ON COLUMN sso.users."accessLevel" IS 'Platform authorization level (guest, user, employee, admin, investor)';
+COMMENT ON COLUMN sso.users.profiles IS 'Business identity types (client, fan, escort, camgirl, etc.) - users can have multiple';
+COMMENT ON COLUMN sso.users."primaryProfile" IS 'Primary business profile for display purposes';