diff --git a/backend/evaluation_helper.py b/backend/evaluation_helper.py
new file mode 100644
index 0000000..f622ca2
--- /dev/null
+++ b/backend/evaluation_helper.py
@@ -0,0 +1,252 @@
+"""
+Training Type Profiles - Helper Functions
+Utilities for loading parameters, profiles, and running evaluations.
+
+Issue: #15
+Date: 2026-03-23
+"""
+from typing import Dict, Optional, List
+import logging
+
+from db import get_cursor
+from profile_evaluator import TrainingProfileEvaluator
+
+logger = logging.getLogger(__name__)
+
+
+def load_parameters_registry(cur) -> Dict[str, Dict]:
+    """
+    Loads training parameters registry from database.
+
+    Returns:
+        Dict mapping parameter_key -> config
+    """
+    cur.execute("""
+        SELECT key, name_de, name_en, category, data_type, unit,
+               description_de, source_field, validation_rules
+        FROM training_parameters
+        WHERE is_active = true
+    """)
+
+    registry = {}
+    for row in cur.fetchall():
+        registry[row['key']] = dict(row)
+
+    return registry
+
+
+def load_training_type_profile(cur, training_type_id: int) -> Optional[Dict]:
+    """
+    Loads training type profile for a given type ID.
+
+    Returns:
+        Profile JSONB or None if not configured
+    """
+    cur.execute(
+        "SELECT profile FROM training_types WHERE id = %s",
+        (training_type_id,)
+    )
+    row = cur.fetchone()
+
+    if row and row['profile']:
+        return row['profile']
+
+    return None
+
+
+def load_evaluation_context(
+    cur,
+    profile_id: str,
+    activity_date: str,
+    lookback_days: int = 30
+) -> Dict:
+    """
+    Loads context data for evaluation (user profile + recent activities).
+
+    Args:
+        cur: Database cursor
+        profile_id: User profile ID
+        activity_date: Date of activity being evaluated
+        lookback_days: How many days of history to load
+
+    Returns:
+        {
+            "user_profile": {...},
+            "recent_activities": [...],
+            "historical_activities": [...]
+        }
+    """
+    # Load user profile
+    cur.execute(
+        "SELECT hf_max, sleep_goal_minutes FROM profiles WHERE id = %s",
+        (profile_id,)
+    )
+    user_row = cur.fetchone()
+    user_profile = dict(user_row) if user_row else {}
+
+    # Load recent activities (last N days)
+    cur.execute("""
+        SELECT id, date, training_type_id, duration_min, hr_avg, hr_max,
+               distance_km, kcal_active, rpe
+        FROM activity_log
+        WHERE profile_id = %s
+          AND date >= %s::date - INTERVAL '%s days'
+          AND date < %s::date
+        ORDER BY date DESC
+        LIMIT 50
+    """, (profile_id, activity_date, lookback_days, activity_date))
+
+    recent_activities = [dict(r) for r in cur.fetchall()]
+
+    # Historical activities (same for MVP)
+    historical_activities = recent_activities
+
+    return {
+        "user_profile": user_profile,
+        "recent_activities": recent_activities,
+        "historical_activities": historical_activities
+    }
+
+
+def evaluate_and_save_activity(
+    cur,
+    activity_id: str,
+    activity_data: Dict,
+    training_type_id: int,
+    profile_id: str
+) -> Optional[Dict]:
+    """
+    Evaluates an activity and saves the result to the database.
+
+    Args:
+        cur: Database cursor
+        activity_id: Activity ID
+        activity_data: Activity data dict
+        training_type_id: Training type ID
+        profile_id: User profile ID
+
+    Returns:
+        Evaluation result or None if no profile configured
+    """
+    # Load profile
+    profile = load_training_type_profile(cur, training_type_id)
+    if not profile:
+        logger.info(f"[EVALUATION] No profile for training_type {training_type_id}, skipping")
+        return None
+
+    # Load parameters registry
+    parameters = load_parameters_registry(cur)
+
+    # Load context
+    context = load_evaluation_context(
+        cur,
+        profile_id,
+        activity_data.get("date"),
+        lookback_days=30
+    )
+
+    # Evaluate
+    evaluator = TrainingProfileEvaluator(parameters)
+    evaluation_result = evaluator.evaluate_activity(
+        activity_data,
+        profile,
+        context
+    )
+
+    # Save to database
+    from psycopg2.extras import Json
+
+    cur.execute("""
+        UPDATE activity_log
+        SET evaluation = %s,
+            quality_label = %s,
+            overall_score = %s
+        WHERE id = %s
+    """, (
+        Json(evaluation_result),
+        evaluation_result.get("quality_label"),
+        evaluation_result.get("overall_score"),
+        activity_id
+    ))
+
+    logger.info(
+        f"[EVALUATION] Activity {activity_id}: "
+        f"{evaluation_result.get('quality_label')} "
+        f"(score: {evaluation_result.get('overall_score')})"
+    )
+
+    return evaluation_result
+
+
+def batch_evaluate_activities(
+    cur,
+    profile_id: str,
+    limit: Optional[int] = None
+) -> Dict:
+    """
+    Re-evaluates all activities for a user.
+
+    Useful for:
+    - Initial setup after profiles are configured
+    - Re-evaluation after profile changes
+
+    Args:
+        cur: Database cursor
+        profile_id: User profile ID
+        limit: Optional limit for testing
+
+    Returns:
+        {
+            "total": int,
+            "evaluated": int,
+            "skipped": int,
+            "errors": int
+        }
+    """
+    # Load all activities
+    query = """
+        SELECT id, profile_id, date, training_type_id, duration_min,
+               hr_avg, hr_max, distance_km, kcal_active, kcal_resting,
+               rpe, pace_min_per_km, cadence, elevation_gain
+        FROM activity_log
+        WHERE profile_id = %s
+        ORDER BY date DESC
+    """
+    params = [profile_id]
+
+    if limit:
+        query += " LIMIT %s"
+        params.append(limit)
+
+    cur.execute(query, params)
+    activities = cur.fetchall()
+
+    stats = {
+        "total": len(activities),
+        "evaluated": 0,
+        "skipped": 0,
+        "errors": 0
+    }
+
+    for activity in activities:
+        activity_dict = dict(activity)
+        try:
+            result = evaluate_and_save_activity(
+                cur,
+                activity_dict["id"],
+                activity_dict,
+                activity_dict["training_type_id"],
+                profile_id
+            )
+
+            if result:
+                stats["evaluated"] += 1
+            else:
+                stats["skipped"] += 1
+
+        except Exception as e:
+            logger.error(f"[BATCH-EVAL] Error evaluating {activity_dict['id']}: {e}")
+            stats["errors"] += 1
+
+    logger.info(f"[BATCH-EVAL] Completed: {stats}")
+    return stats
diff --git a/backend/main.py b/backend/main.py
index fba2846..a7689ec 100644
--- a/backend/main.py
+++ b/backend/main.py
@@ -21,6 +21,7 @@ from routers import admin, stats, exportdata, importdata
 from routers import subscription, coupons, features, tiers_mgmt, tier_limits
 from routers import user_restrictions, access_grants, training_types, admin_training_types
 from routers import admin_activity_mappings, sleep, rest_days
+from routers import evaluation  # v9d/v9e Training Type Profiles (#15)
 
 # ── App Configuration ─────────────────────────────────────────────────────────
 DATA_DIR   = Path(os.getenv("DATA_DIR",   "./data"))
@@ -92,6 +93,7 @@ app.include_router(admin_training_types.router) # /api/admin/training-types/*
 app.include_router(admin_activity_mappings.router) # /api/admin/activity-mappings/*
 app.include_router(sleep.router)                # /api/sleep/* (v9d Phase 2b)
 app.include_router(rest_days.router)            # /api/rest-days/* (v9d Phase 2a)
+app.include_router(evaluation.router)           # /api/evaluation/* (v9d/v9e Training Profiles #15)
 
 # ── Health Check ──────────────────────────────────────────────────────────────
 @app.get("/")
diff --git a/backend/migrations/013_training_parameters.sql b/backend/migrations/013_training_parameters.sql
new file mode 100644
index 0000000..9f80bcf
--- /dev/null
+++ b/backend/migrations/013_training_parameters.sql
@@ -0,0 +1,144 @@
+-- Migration 013: Training Parameters Registry
+-- Training Type Profiles System - Foundation
+-- Date: 2026-03-23
+-- Issue: #15
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- TRAINING PARAMETERS REGISTRY
+-- Zentrale Definition aller messbaren Parameter für Aktivitäten
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+CREATE TABLE IF NOT EXISTS training_parameters (
+    id SERIAL PRIMARY KEY,
+    key VARCHAR(50) UNIQUE NOT NULL,
+    name_de VARCHAR(100) NOT NULL,
+    name_en VARCHAR(100) NOT NULL,
+    category VARCHAR(50) NOT NULL,
+    data_type VARCHAR(20) NOT NULL,
+    unit VARCHAR(20),
+    description_de TEXT,
+    description_en TEXT,
+    source_field VARCHAR(100),
+    validation_rules JSONB DEFAULT '{}'::jsonb,
+    is_active BOOLEAN DEFAULT true,
+    created_at TIMESTAMP DEFAULT NOW(),
+
+    CONSTRAINT chk_category CHECK (category IN (
+        'physical', 'physiological', 'subjective', 'environmental', 'performance'
+    )),
+    CONSTRAINT chk_data_type CHECK (data_type IN (
+        'integer', 'float', 'string', 'boolean'
+    ))
+);
+
+CREATE INDEX idx_training_parameters_category ON training_parameters(category) WHERE is_active = true;
+CREATE INDEX idx_training_parameters_key ON training_parameters(key) WHERE is_active = true;
+
+COMMENT ON TABLE training_parameters IS 'Registry of all measurable activity parameters (Training Type Profiles System)';
+COMMENT ON COLUMN training_parameters.key IS 'Unique identifier (e.g. "avg_hr", "duration_min")';
+COMMENT ON COLUMN training_parameters.category IS 'Parameter category: physical, physiological, subjective, environmental, performance';
+COMMENT ON COLUMN training_parameters.data_type IS 'Data type: integer, float, string, boolean';
+COMMENT ON COLUMN training_parameters.source_field IS 'Mapping to activity_log column name';
+COMMENT ON COLUMN training_parameters.validation_rules IS 'Min/Max/Enum for validation (JSONB)';
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- STANDARD PARAMETERS
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+INSERT INTO training_parameters (key, name_de, name_en, category, data_type, unit, source_field, validation_rules, description_de, description_en) VALUES
+
+-- Physical Parameters
+('duration_min', 'Dauer', 'Duration', 'physical', 'integer', 'min', 'duration_min',
+ '{"min": 0, "max": 600}'::jsonb,
+ 'Trainingsdauer in Minuten',
+ 'Training duration in minutes'),
+
+('distance_km', 'Distanz', 'Distance', 'physical', 'float', 'km', 'distance_km',
+ '{"min": 0, "max": 200}'::jsonb,
+ 'Zurückgelegte Distanz in Kilometern',
+ 'Distance covered in kilometers'),
+
+('kcal_active', 'Aktive Kalorien', 'Active Calories', 'physical', 'integer', 'kcal', 'kcal_active',
+ '{"min": 0, "max": 5000}'::jsonb,
+ 'Aktiver Kalorienverbrauch',
+ 'Active calorie burn'),
+
+('kcal_resting', 'Ruhekalorien', 'Resting Calories', 'physical', 'integer', 'kcal', 'kcal_resting',
+ '{"min": 0, "max": 2000}'::jsonb,
+ 'Ruheumsatz während Training',
+ 'Resting calorie burn during training'),
+
+('elevation_gain', 'Höhenmeter', 'Elevation Gain', 'physical', 'integer', 'm', 'elevation_gain',
+ '{"min": 0, "max": 5000}'::jsonb,
+ 'Überwundene Höhenmeter',
+ 'Elevation gain in meters'),
+
+('pace_min_per_km', 'Pace', 'Pace', 'physical', 'float', 'min/km', 'pace_min_per_km',
+ '{"min": 2, "max": 20}'::jsonb,
+ 'Durchschnittstempo in Minuten pro Kilometer',
+ 'Average pace in minutes per kilometer'),
+
+('cadence', 'Trittfrequenz', 'Cadence', 'physical', 'integer', 'spm', 'cadence',
+ '{"min": 0, "max": 220}'::jsonb,
+ 'Schrittfrequenz (Schritte pro Minute)',
+ 'Step frequency (steps per minute)'),
+
+-- Physiological Parameters
+('avg_hr', 'Durchschnittspuls', 'Average Heart Rate', 'physiological', 'integer', 'bpm', 'hr_avg',
+ '{"min": 30, "max": 220}'::jsonb,
+ 'Durchschnittliche Herzfrequenz',
+ 'Average heart rate'),
+
+('max_hr', 'Maximalpuls', 'Max Heart Rate', 'physiological', 'integer', 'bpm', 'hr_max',
+ '{"min": 40, "max": 220}'::jsonb,
+ 'Maximale Herzfrequenz',
+ 'Maximum heart rate'),
+
+('min_hr', 'Minimalpuls', 'Min Heart Rate', 'physiological', 'integer', 'bpm', 'hr_min',
+ '{"min": 30, "max": 200}'::jsonb,
+ 'Minimale Herzfrequenz',
+ 'Minimum heart rate'),
+
+('avg_power', 'Durchschnittsleistung', 'Average Power', 'physiological', 'integer', 'W', 'avg_power',
+ '{"min": 0, "max": 1000}'::jsonb,
+ 'Durchschnittliche Leistung in Watt',
+ 'Average power output in watts'),
+
+-- Subjective Parameters
+('rpe', 'RPE (Anstrengung)', 'RPE (Perceived Exertion)', 'subjective', 'integer', 'scale', 'rpe',
+ '{"min": 1, "max": 10}'::jsonb,
+ 'Subjektive Anstrengung (Rate of Perceived Exertion)',
+ 'Rate of Perceived Exertion'),
+
+-- Environmental Parameters
+('temperature_celsius', 'Temperatur', 'Temperature', 'environmental', 'float', '°C', 'temperature_celsius',
+ '{"min": -30, "max": 50}'::jsonb,
+ 'Umgebungstemperatur in Celsius',
+ 'Ambient temperature in Celsius'),
+
+('humidity_percent', 'Luftfeuchtigkeit', 'Humidity', 'environmental', 'integer', '%', 'humidity_percent',
+ '{"min": 0, "max": 100}'::jsonb,
+ 'Relative Luftfeuchtigkeit in Prozent',
+ 'Relative humidity in percent'),
+
+-- Performance Parameters (calculated)
+('avg_hr_percent', '% Max-HF', '% Max HR', 'performance', 'float', '%', 'avg_hr_percent',
+ '{"min": 0, "max": 100}'::jsonb,
+ 'Durchschnittspuls als Prozent der maximalen Herzfrequenz',
+ 'Average heart rate as percentage of max heart rate'),
+
+('kcal_per_km', 'Kalorien pro km', 'Calories per km', 'performance', 'float', 'kcal/km', 'kcal_per_km',
+ 'Kalorienverbrauch pro Kilometer',
+ 'Calorie burn per kilometer');
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- SUMMARY
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+-- Display inserted parameters
+DO $$
+BEGIN
+    RAISE NOTICE '✓ Migration 013 completed';
+    RAISE NOTICE '  - Created training_parameters table';
+    RAISE NOTICE '  - Inserted % standard parameters', (SELECT COUNT(*) FROM training_parameters);
+END $$;
diff --git a/backend/migrations/014_training_profiles.sql b/backend/migrations/014_training_profiles.sql
new file mode 100644
index 0000000..6747c6c
--- /dev/null
+++ b/backend/migrations/014_training_profiles.sql
@@ -0,0 +1,114 @@
+-- Migration 014: Training Type Profiles & Activity Evaluation
+-- Training Type Profiles System - Schema Extensions
+-- Date: 2026-03-23
+-- Issue: #15
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- EXTEND TRAINING TYPES
+-- Add profile column for comprehensive training type configuration
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+ALTER TABLE training_types ADD COLUMN IF NOT EXISTS profile JSONB DEFAULT NULL;
+
+CREATE INDEX idx_training_types_profile_enabled ON training_types
+    ((profile->'rule_sets'->'minimum_requirements'->>'enabled'))
+    WHERE profile IS NOT NULL;
+
+COMMENT ON COLUMN training_types.profile IS 'Comprehensive training type profile with 7 dimensions (rule_sets, intensity_zones, training_effects, periodization, performance_indicators, safety, ai_context)';
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- EXTEND ACTIVITY LOG
+-- Add evaluation results and quality labels
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS evaluation JSONB DEFAULT NULL;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS quality_label VARCHAR(20);
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS overall_score FLOAT;
+
+CREATE INDEX idx_activity_quality_label ON activity_log(quality_label)
+    WHERE quality_label IS NOT NULL;
+
+CREATE INDEX idx_activity_overall_score ON activity_log(overall_score DESC)
+    WHERE overall_score IS NOT NULL;
+
+CREATE INDEX idx_activity_evaluation_passed ON activity_log
+    ((evaluation->'rule_set_results'->'minimum_requirements'->>'passed'))
+    WHERE evaluation IS NOT NULL;
+
+COMMENT ON COLUMN activity_log.evaluation IS 'Complete evaluation result (7 dimensions, scores, recommendations, warnings)';
+COMMENT ON COLUMN activity_log.quality_label IS 'Quality label: excellent, good, acceptable, poor (for quick filtering)';
+COMMENT ON COLUMN activity_log.overall_score IS 'Overall quality score 0.0-1.0 (for sorting)';
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- ADD MISSING COLUMNS (if not already added by previous migrations)
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+-- Add HR columns if not exist (might be in Migration 008)
+DO $$
+BEGIN
+    IF NOT EXISTS (SELECT 1 FROM information_schema.columns
+                   WHERE table_name='activity_log' AND column_name='hr_min') THEN
+        ALTER TABLE activity_log ADD COLUMN hr_min INTEGER CHECK (hr_min > 0 AND hr_min < 200);
+    END IF;
+END $$;
+
+-- Add performance columns for calculated values
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS avg_hr_percent FLOAT;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS kcal_per_km FLOAT;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS pace_min_per_km FLOAT;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS cadence INTEGER;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS avg_power INTEGER;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS elevation_gain INTEGER;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS temperature_celsius FLOAT;
+ALTER TABLE activity_log ADD COLUMN IF NOT EXISTS humidity_percent INTEGER;
+
+COMMENT ON COLUMN activity_log.avg_hr_percent IS 'Average HR as percentage of user max HR (calculated)';
+COMMENT ON COLUMN activity_log.kcal_per_km IS 'Calories burned per kilometer (calculated)';
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- HELPER FUNCTION: Calculate avg_hr_percent
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+CREATE OR REPLACE FUNCTION calculate_avg_hr_percent()
+RETURNS TRIGGER AS $$
+DECLARE
+    user_max_hr INTEGER;
+BEGIN
+    -- Get user's max HR from profile
+    SELECT hf_max INTO user_max_hr
+    FROM profiles
+    WHERE id = NEW.profile_id;
+
+    -- Calculate percentage if both values exist
+    IF NEW.hr_avg IS NOT NULL AND user_max_hr IS NOT NULL AND user_max_hr > 0 THEN
+        NEW.avg_hr_percent := (NEW.hr_avg::float / user_max_hr::float) * 100;
+    END IF;
+
+    -- Calculate kcal per km
+    IF NEW.kcal_active IS NOT NULL AND NEW.distance_km IS NOT NULL AND NEW.distance_km > 0 THEN
+        NEW.kcal_per_km := NEW.kcal_active::float / NEW.distance_km;
+    END IF;
+
+    RETURN NEW;
+END;
+$$ LANGUAGE plpgsql;
+
+-- Trigger for automatic calculation
+DROP TRIGGER IF EXISTS trg_calculate_performance_metrics ON activity_log;
+CREATE TRIGGER trg_calculate_performance_metrics
+    BEFORE INSERT OR UPDATE ON activity_log
+    FOR EACH ROW
+    EXECUTE FUNCTION calculate_avg_hr_percent();
+
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+-- SUMMARY
+-- ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+DO $$
+BEGIN
+    RAISE NOTICE '✓ Migration 014 completed';
+    RAISE NOTICE '  - Extended training_types with profile column';
+    RAISE NOTICE '  - Extended activity_log with evaluation columns';
+    RAISE NOTICE '  - Added performance metric calculations';
+    RAISE NOTICE '  - Created indexes for fast queries';
+END $$;
diff --git a/backend/profile_evaluator.py b/backend/profile_evaluator.py
new file mode 100644
index 0000000..c2f05a3
--- /dev/null
+++ b/backend/profile_evaluator.py
@@ -0,0 +1,349 @@
+"""
+Training Type Profiles - Master Evaluator
+Comprehensive activity evaluation across all 7 dimensions.
+
+Issue: #15
+Date: 2026-03-23
+"""
+from typing import Dict, Optional, List
+from datetime import datetime
+import logging
+
+from rule_engine import RuleEvaluator, IntensityZoneEvaluator, TrainingEffectsEvaluator
+
+logger = logging.getLogger(__name__)
+
+
+class TrainingProfileEvaluator:
+    """
+    Master class for comprehensive activity evaluation.
+
+    Evaluates an activity against a training type profile across 7 dimensions:
+    1. Minimum Requirements (Quality Gates)
+    2. Intensity Zones (HR zones)
+    3. Training Effects (Abilities)
+    4. Periodization (Frequency & Recovery)
+    5. Performance Indicators (KPIs)
+    6. Safety (Warnings)
+    7. AI Context
+    """
+
+    def __init__(self, parameters_registry: Dict[str, Dict]):
+        """
+        Initialize evaluator with parameter registry.
+
+        Args:
+            parameters_registry: Dict mapping parameter_key -> config
+        """
+        self.parameters_registry = parameters_registry
+        self.rule_evaluator = RuleEvaluator()
+        self.zone_evaluator = IntensityZoneEvaluator()
+        self.effects_evaluator = TrainingEffectsEvaluator()
+
+    def evaluate_activity(
+        self,
+        activity: Dict,
+        training_type_profile: Optional[Dict],
+        context: Optional[Dict] = None
+    ) -> Dict:
+        """
+        Complete evaluation of an activity against its training type profile.
+
+        Args:
+            activity: Activity data dictionary
+            training_type_profile: Training type profile (JSONB)
+            context: {
+                "user_profile": {...},
+                "recent_activities": [...],
+                "historical_activities": [...]
+            }
+
+        Returns:
+            {
+                "evaluated_at": ISO timestamp,
+                "profile_version": str,
+                "rule_set_results": {
+                    "minimum_requirements": {...},
+                    "intensity_zones": {...},
+                    "training_effects": {...},
+                    "periodization": {...},
+                    "performance_indicators": {...},
+                    "safety": {...}
+                },
+                "overall_score": float (0-1),
+                "quality_label": str,
+                "recommendations": [str],
+                "warnings": [str]
+            }
+        """
+        # No profile? Return unvalidated result
+        if not training_type_profile:
+            return self._create_unvalidated_result()
+
+        rule_sets = training_type_profile.get("rule_sets", {})
+        context = context or {}
+
+        results = {
+            "evaluated_at": datetime.now().isoformat(),
+            "profile_version": training_type_profile.get("version", "unknown"),
+            "rule_set_results": {}
+        }
+
+        # ━━━ 1. MINIMUM REQUIREMENTS ━━━
+        if "minimum_requirements" in rule_sets:
+            results["rule_set_results"]["minimum_requirements"] = \
+                self.rule_evaluator.evaluate_rule_set(
+                    rule_sets["minimum_requirements"],
+                    activity,
+                    self.parameters_registry
+                )
+
+        # ━━━ 2. INTENSITY ZONES ━━━
+        if "intensity_zones" in rule_sets:
+            results["rule_set_results"]["intensity_zones"] = \
+                self.zone_evaluator.evaluate(
+                    rule_sets["intensity_zones"],
+                    activity,
+                    context.get("user_profile", {})
+                )
+
+        # ━━━ 3. TRAINING EFFECTS ━━━
+        if "training_effects" in rule_sets:
+            results["rule_set_results"]["training_effects"] = \
+                self.effects_evaluator.evaluate(
+                    rule_sets["training_effects"],
+                    activity,
+                    results["rule_set_results"].get("intensity_zones")
+                )
+
+        # ━━━ 4. PERIODIZATION ━━━
+        if "periodization" in rule_sets:
+            results["rule_set_results"]["periodization"] = \
+                self._evaluate_periodization(
+                    rule_sets["periodization"],
+                    activity,
+                    context.get("recent_activities", [])
+                )
+
+        # ━━━ 5. PERFORMANCE INDICATORS ━━━
+        if "performance_indicators" in rule_sets:
+            results["rule_set_results"]["performance_indicators"] = \
+                self._evaluate_performance(
+                    rule_sets["performance_indicators"],
+                    activity,
+                    context.get("historical_activities", [])
+                )
+
+        # ━━━ 6. SAFETY WARNINGS ━━━
+        if "safety" in rule_sets:
+            results["rule_set_results"]["safety"] = \
+                self._evaluate_safety(
+                    rule_sets["safety"],
+                    activity
+                )
+
+        # ━━━ OVERALL SCORE & QUALITY LABEL ━━━
+        overall_score = self._calculate_overall_score(results["rule_set_results"])
+        results["overall_score"] = overall_score
+        results["quality_label"] = self._get_quality_label(overall_score)
+
+        # ━━━ RECOMMENDATIONS & WARNINGS ━━━
+        results["recommendations"] = self._generate_recommendations(results)
+        results["warnings"] = self._collect_warnings(results)
+
+        return results
+
+    def _create_unvalidated_result(self) -> Dict:
+        """Creates result for activities without profile."""
+        return {
+            "evaluated_at": datetime.now().isoformat(),
+            "profile_version": None,
+            "rule_set_results": {},
+            "overall_score": None,
+            "quality_label": None,
+            "recommendations": ["Kein Trainingsprofil konfiguriert"],
+            "warnings": []
+        }
+
+    def _evaluate_periodization(
+        self,
+        config: Dict,
+        activity: Dict,
+        recent_activities: List[Dict]
+    ) -> Dict:
+        """
+        Evaluates periodization compliance (frequency & recovery).
+
+        Simplified for MVP - full implementation later.
+        """
+        if not config.get("enabled", False):
+            return {"enabled": False}
+
+        # Basic frequency check
+        training_type_id = activity.get("training_type_id")
+        same_type_this_week = sum(
+            1 for a in recent_activities
+            if a.get("training_type_id") == training_type_id
+        )
+
+        frequency_config = config.get("frequency", {})
+        optimal = frequency_config.get("per_week_optimal", 3)
+
+        return {
+            "enabled": True,
+            "weekly_count": same_type_this_week,
+            "optimal_count": optimal,
+            "frequency_status": "optimal" if same_type_this_week <= optimal else "over_optimal",
+            "recovery_adequate": True,  # Simplified for MVP
+            "warning": None
+        }
+
+    def _evaluate_performance(
+        self,
+        config: Dict,
+        activity: Dict,
+        historical_activities: List[Dict]
+    ) -> Dict:
+        """
+        Evaluates performance development.
+
+        Simplified for MVP - full implementation later.
+        """
+        if not config.get("enabled", False):
+            return {"enabled": False}
+
+        return {
+            "enabled": True,
+            "trend": "stable",  # Simplified
+            "metrics_comparison": {},
+            "benchmark_level": "intermediate"
+        }
+
+    def _evaluate_safety(self, config: Dict, activity: Dict) -> Dict:
+        """
+        Evaluates safety warnings.
+        """
+        if not config.get("enabled", False):
+            return {"enabled": False, "warnings": []}
+
+        warnings_config = config.get("warnings", [])
+        triggered_warnings = []
+
+        for warning_rule in warnings_config:
+            param_key = warning_rule.get("parameter")
+            operator = warning_rule.get("operator")
+            threshold = warning_rule.get("value")
+            severity = warning_rule.get("severity", "medium")
+            message = warning_rule.get("message", "")
+
+            actual_value = activity.get(param_key)
+
+            if actual_value is not None:
+                operator_func = RuleEvaluator.OPERATORS.get(operator)
+                if operator_func and operator_func(actual_value, threshold):
+                    triggered_warnings.append({
+                        "severity": severity,
+                        "message": message,
+                        "parameter": param_key,
+                        "actual_value": actual_value,
+                        "threshold": threshold
+                    })
+
+        return {
+            "enabled": True,
+            "warnings": triggered_warnings
+        }
+
+    def _calculate_overall_score(self, rule_set_results: Dict) -> float:
+        """
+        Calculates weighted overall score.
+
+        Weights:
+        - Minimum Requirements: 40%
+        - Intensity Zones: 20%
+        - Periodization: 20%
+        - Performance: 10%
+        - Training Effects: 10%
+        """
+        weights = {
+            "minimum_requirements": 0.4,
+            "intensity_zones": 0.2,
+            "periodization": 0.2,
+            "performance_indicators": 0.1,
+            "training_effects": 0.1
+        }
+
+        total_score = 0.0
+        total_weight = 0.0
+
+        for rule_set_name, weight in weights.items():
+            result = rule_set_results.get(rule_set_name)
+            if result and result.get("enabled"):
+                score = result.get("score", 0.5)
+
+                # Special handling for different result types
+                if rule_set_name == "intensity_zones":
+                    score = result.get("duration_quality", 0.5)
+                elif rule_set_name == "periodization":
+                    score = 1.0 if result.get("recovery_adequate", False) else 0.5
+
+                total_score += score * weight
+                total_weight += weight
+
+        return round(total_score / total_weight, 2) if total_weight > 0 else 0.5
+
+    def _get_quality_label(self, score: Optional[float]) -> Optional[str]:
+        """Converts score to quality label."""
+        if score is None:
+            return None
+
+        if score >= 0.9:
+            return "excellent"
+        elif score >= 0.7:
+            return "good"
+        elif score >= 0.5:
+            return "acceptable"
+        else:
+            return "poor"
+
+    def _generate_recommendations(self, results: Dict) -> List[str]:
+        """Generates actionable recommendations."""
+        recommendations = []
+
+        # Check minimum requirements
+        min_req = results["rule_set_results"].get("minimum_requirements", {})
+        if min_req.get("enabled") and not min_req.get("passed"):
+            for failed in min_req.get("failed_rules", []):
+                param = failed.get("parameter")
+                actual = failed.get("actual_value")
+                expected = failed.get("expected_value")
+                reason = failed.get("reason", "")
+                symbol = failed.get("operator_symbol", "")
+
+                recommendations.append(
+                    f"{param}: {actual} {symbol} {expected} - {reason}"
+                )
+
+        # Check intensity zones
+        zone_result = results["rule_set_results"].get("intensity_zones", {})
+        if zone_result.get("enabled") and zone_result.get("recommendation"):
+            recommendations.append(zone_result["recommendation"])
+
+        # Default recommendation if excellent
+        if results.get("quality_label") == "excellent" and not recommendations:
+            recommendations.append("Hervorragendes Training! Weiter so.")
+
+        return recommendations
+
+    def _collect_warnings(self, results: Dict) -> List[str]:
+        """Collects all warnings from safety checks."""
+        safety_result = results["rule_set_results"].get("safety", {})
+        if not safety_result.get("enabled"):
+            return []
+
+        warnings = []
+        for warning in safety_result.get("warnings", []):
+            severity_icon = "🔴" if warning["severity"] == "high" else "⚠️"
+            warnings.append(f"{severity_icon} {warning['message']}")
+
+        return warnings
diff --git a/backend/routers/evaluation.py b/backend/routers/evaluation.py
new file mode 100644
index 0000000..bc2370f
--- /dev/null
+++ b/backend/routers/evaluation.py
@@ -0,0 +1,146 @@
+"""
+Evaluation Endpoints - Training Type Profiles
+Endpoints for activity evaluation and re-evaluation.
+
+Issue: #15
+Date: 2026-03-23
+"""
+import logging
+from typing import Optional
+from fastapi import APIRouter, HTTPException, Depends
+
+from db import get_db, get_cursor, r2d
+from auth import require_auth, require_admin
+from evaluation_helper import (
+    evaluate_and_save_activity,
+    batch_evaluate_activities,
+    load_parameters_registry
+)
+
+router = APIRouter(prefix="/api/evaluation", tags=["evaluation"])
+logger = logging.getLogger(__name__)
+
+
+@router.get("/parameters")
+def list_parameters(session: dict = Depends(require_auth)):
+    """
+    List all available training parameters.
+    """
+    with get_db() as conn:
+        cur = get_cursor(conn)
+        parameters = load_parameters_registry(cur)
+
+    return {
+        "parameters": list(parameters.values()),
+        "count": len(parameters)
+    }
+
+
+@router.post("/activity/{activity_id}")
+def evaluate_activity(
+    activity_id: str,
+    session: dict = Depends(require_auth)
+):
+    """
+    Evaluates or re-evaluates a single activity.
+
+    Returns the evaluation result.
+    """
+    profile_id = session['profile_id']
+
+    with get_db() as conn:
+        cur = get_cursor(conn)
+
+        # Load activity
+        cur.execute("""
+            SELECT id, profile_id, date, training_type_id, duration_min,
+                   hr_avg, hr_max, distance_km, kcal_active, kcal_resting,
+                   rpe, pace_min_per_km, cadence, elevation_gain
+            FROM activity_log
+            WHERE id = %s AND profile_id = %s
+        """, (activity_id, profile_id))
+
+        activity = cur.fetchone()
+        if not activity:
+            raise HTTPException(404, "Activity not found")
+
+        activity_dict = dict(activity)
+
+        # Evaluate
+        result = evaluate_and_save_activity(
+            cur,
+            activity_dict["id"],
+            activity_dict,
+            activity_dict["training_type_id"],
+            profile_id
+        )
+
+        if not result:
+            return {
+                "message": "No profile configured for this training type",
+                "evaluation": None
+            }
+
+    return {
+        "message": "Activity evaluated",
+        "evaluation": result
+    }
+
+
+@router.post("/batch")
+def batch_evaluate(
+    limit: Optional[int] = None,
+    session: dict = Depends(require_auth)
+):
+    """
+    Re-evaluates all activities for the current user.
+
+    Optional limit parameter for testing.
+    """
+    profile_id = session['profile_id']
+
+    with get_db() as conn:
+        cur = get_cursor(conn)
+        stats = batch_evaluate_activities(cur, profile_id, limit)
+
+    return {
+        "message": "Batch evaluation completed",
+        "stats": stats
+    }
+
+
+@router.post("/batch/all")
+def batch_evaluate_all(session: dict = Depends(require_admin)):
+    """
+    Admin-only: Re-evaluates all activities for all users.
+
+    Use with caution on large databases!
+    """
+    with get_db() as conn:
+        cur = get_cursor(conn)
+
+        # Get all profiles
+        cur.execute("SELECT id FROM profiles")
+        profiles = cur.fetchall()
+
+        total_stats = {
+            "profiles": len(profiles),
+            "total": 0,
+            "evaluated": 0,
+            "skipped": 0,
+            "errors": 0
+        }
+
+        for profile in profiles:
+            profile_id = profile['id']
+            stats = batch_evaluate_activities(cur, profile_id)
+
+            total_stats["total"] += stats["total"]
+            total_stats["evaluated"] += stats["evaluated"]
+            total_stats["skipped"] += stats["skipped"]
+            total_stats["errors"] += stats["errors"]
+
+    return {
+        "message": "Batch evaluation for all users completed",
+        "stats": total_stats
+    }
diff --git a/backend/rule_engine.py b/backend/rule_engine.py
new file mode 100644
index 0000000..e0cff89
--- /dev/null
+++ b/backend/rule_engine.py
@@ -0,0 +1,427 @@
+"""
+Training Type Profiles - Rule Engine
+Flexible rule evaluation system for activity quality assessment.
+
+Issue: #15
+Date: 2026-03-23
+"""
+from typing import Any, Dict, List, Optional, Callable
+from datetime import datetime
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class RuleEvaluator:
+    """
+    Generic rule evaluator for arbitrary parameters and operators.
+
+    Supports flexible rule definitions with various operators:
+    - gte, lte, gt, lt: Comparison operators
+    - eq, neq: Equality operators
+    - between: Range checks
+    - in, not_in: Set membership
+    """
+
+    # Operator definitions
+    OPERATORS: Dict[str, Callable[[Any, Any], bool]] = {
+        "gte": lambda actual, expected: actual is not None and actual >= expected,
+        "lte": lambda actual, expected: actual is not None and actual <= expected,
+        "gt": lambda actual, expected: actual is not None and actual > expected,
+        "lt": lambda actual, expected: actual is not None and actual < expected,
+        "eq": lambda actual, expected: actual == expected,
+        "neq": lambda actual, expected: actual != expected,
+        "between": lambda actual, expected: actual is not None and expected[0] <= actual <= expected[1],
+        "in": lambda actual, expected: actual in expected,
+        "not_in": lambda actual, expected: actual not in expected,
+    }
+
+    OPERATOR_SYMBOLS = {
+        "gte": "≥",
+        "lte": "≤",
+        "gt": ">",
+        "lt": "<",
+        "eq": "=",
+        "neq": "≠",
+        "between": "⟷",
+        "in": "∈",
+        "not_in": "∉",
+    }
+
+    @classmethod
+    def evaluate_rule(
+        cls,
+        rule: Dict,
+        activity: Dict,
+        parameters_registry: Dict[str, Dict]
+    ) -> Dict:
+        """
+        Evaluates a single rule against an activity.
+
+        Args:
+            rule: {
+                "parameter": str,
+                "operator": str,
+                "value": Any,
+                "weight": int,
+                "optional": bool,
+                "reason": str
+            }
+            activity: Activity data dictionary
+            parameters_registry: Mapping parameter_key -> config
+
+        Returns:
+            {
+                "passed": bool,
+                "actual_value": Any,
+                "expected_value": Any,
+                "parameter": str,
+                "operator": str,
+                "operator_symbol": str,
+                "reason": str,
+                "weight": int,
+                "skipped": bool (optional),
+                "error": str (optional)
+            }
+        """
+        param_key = rule.get("parameter")
+        operator = rule.get("operator")
+        expected_value = rule.get("value")
+        weight = rule.get("weight", 1)
+        reason = rule.get("reason", "")
+        optional = rule.get("optional", False)
+
+        # Get parameter configuration
+        param_config = parameters_registry.get(param_key)
+        if not param_config:
+            return {
+                "passed": False,
+                "parameter": param_key,
+                "error": f"Unknown parameter: {param_key}"
+            }
+
+        # Extract value from activity
+        source_field = param_config.get("source_field", param_key)
+        actual_value = activity.get(source_field)
+
+        # Optional and not provided? → Pass
+        if optional and actual_value is None:
+            return {
+                "passed": True,
+                "actual_value": None,
+                "expected_value": expected_value,
+                "parameter": param_key,
+                "operator": operator,
+                "operator_symbol": cls.OPERATOR_SYMBOLS.get(operator, operator),
+                "reason": "Optional parameter not provided",
+                "weight": weight,
+                "skipped": True
+            }
+
+        # Required but not provided? → Fail
+        if actual_value is None:
+            return {
+                "passed": False,
+                "actual_value": None,
+                "expected_value": expected_value,
+                "parameter": param_key,
+                "operator": operator,
+                "operator_symbol": cls.OPERATOR_SYMBOLS.get(operator, operator),
+                "reason": reason or "Required parameter missing",
+                "weight": weight
+            }
+
+        # Apply operator
+        operator_func = cls.OPERATORS.get(operator)
+        if not operator_func:
+            return {
+                "passed": False,
+                "parameter": param_key,
+                "error": f"Unknown operator: {operator}"
+            }
+
+        try:
+            passed = operator_func(actual_value, expected_value)
+        except Exception as e:
+            logger.error(f"[RULE-ENGINE] Error evaluating rule {param_key}: {e}")
+            return {
+                "passed": False,
+                "parameter": param_key,
+                "error": f"Evaluation error: {str(e)}"
+            }
+
+        return {
+            "passed": passed,
+            "actual_value": actual_value,
+            "expected_value": expected_value,
+            "parameter": param_key,
+            "operator": operator,
+            "operator_symbol": cls.OPERATOR_SYMBOLS.get(operator, operator),
+            "reason": reason,
+            "weight": weight
+        }
+
+    @classmethod
+    def evaluate_rule_set(
+        cls,
+        rule_set: Dict,
+        activity: Dict,
+        parameters_registry: Dict[str, Dict]
+    ) -> Dict:
+        """
+        Evaluates a complete rule set (e.g., minimum_requirements).
+
+        Args:
+            rule_set: {
+                "enabled": bool,
+                "pass_strategy": str,
+                "pass_threshold": float,
+                "rules": [...]
+            }
+            activity: Activity data
+            parameters_registry: Parameter configurations
+
+        Returns:
+            {
+                "enabled": bool,
+                "passed": bool,
+                "score": float (0-1),
+                "rule_results": [...],
+                "pass_strategy": str,
+                "pass_threshold": float,
+                "failed_rules": [...]
+            }
+        """
+        if not rule_set.get("enabled", False):
+            return {
+                "enabled": False,
+                "passed": True,
+                "score": 1.0,
+                "rule_results": [],
+                "failed_rules": []
+            }
+
+        rules = rule_set.get("rules", [])
+        pass_strategy = rule_set.get("pass_strategy", "weighted_score")
+        pass_threshold = rule_set.get("pass_threshold", 0.6)
+
+        rule_results = []
+        failed_rules = []
+        total_weight = 0
+        passed_weight = 0
+
+        # Evaluate each rule
+        for rule in rules:
+            result = cls.evaluate_rule(rule, activity, parameters_registry)
+            rule_results.append(result)
+
+            if result.get("skipped"):
+                continue
+
+            if result.get("error"):
+                logger.warning(f"[RULE-ENGINE] Rule error: {result['error']}")
+                continue
+
+            weight = result.get("weight", 1)
+            total_weight += weight
+
+            if result["passed"]:
+                passed_weight += weight
+            else:
+                failed_rules.append(result)
+
+        # Calculate score
+        score = passed_weight / total_weight if total_weight > 0 else 1.0
+
+        # Apply pass strategy
+        if pass_strategy == "all_must_pass":
+            passed = all(
+                r["passed"] for r in rule_results
+                if not r.get("skipped") and not r.get("error")
+            )
+        elif pass_strategy == "weighted_score":
+            passed = score >= pass_threshold
+        elif pass_strategy == "at_least_n":
+            n = rule_set.get("at_least_n", 1)
+            passed_count = sum(
+                1 for r in rule_results
+                if r["passed"] and not r.get("skipped")
+            )
+            passed = passed_count >= n
+        else:
+            passed = False
+            logger.warning(f"[RULE-ENGINE] Unknown pass strategy: {pass_strategy}")
+
+        return {
+            "enabled": True,
+            "passed": passed,
+            "score": round(score, 2),
+            "rule_results": rule_results,
+            "failed_rules": failed_rules,
+            "pass_strategy": pass_strategy,
+            "pass_threshold": pass_threshold
+        }
+
+
+class IntensityZoneEvaluator:
+    """
+    Evaluates heart rate zones and time distribution.
+    """
+
+    @staticmethod
+    def evaluate(
+        zone_config: Dict,
+        activity: Dict,
+        user_profile: Dict
+    ) -> Dict:
+        """
+        Evaluates which HR zone the activity was in.
+
+        Args:
+            zone_config: intensity_zones configuration
+            activity: Activity data (with hr_avg)
+            user_profile: User profile (with hf_max)
+
+        Returns:
+            {
+                "enabled": bool,
+                "dominant_zone": str,
+                "avg_hr_percent": float,
+                "zone_color": str,
+                "zone_effect": str,
+                "duration_quality": float (0-1),
+                "recommendation": str
+            }
+        """
+        if not zone_config.get("enabled", False):
+            return {"enabled": False}
+
+        avg_hr = activity.get("hr_avg")
+        user_max_hr = user_profile.get("hf_max", 180)  # Default 180 if not set
+
+        if not avg_hr or not user_max_hr:
+            return {
+                "enabled": True,
+                "dominant_zone": "unknown",
+                "avg_hr_percent": None,
+                "recommendation": "Herzfrequenz-Daten fehlen"
+            }
+
+        avg_hr_percent = (avg_hr / user_max_hr) * 100
+
+        # Find matching zone
+        zones = zone_config.get("zones", [])
+        dominant_zone = None
+
+        for zone in zones:
+            zone_rules = zone.get("rules", [])
+            for rule in zone_rules:
+                if rule["parameter"] == "avg_hr_percent":
+                    min_percent, max_percent = rule["value"]
+                    if min_percent <= avg_hr_percent <= max_percent:
+                        dominant_zone = zone
+                        break
+            if dominant_zone:
+                break
+
+        if not dominant_zone:
+            return {
+                "enabled": True,
+                "dominant_zone": "out_of_range",
+                "avg_hr_percent": round(avg_hr_percent, 1),
+                "recommendation": "Herzfrequenz außerhalb definierter Zonen"
+            }
+
+        # Check duration quality
+        duration = activity.get("duration_min", 0)
+        target_duration = dominant_zone.get("target_duration_min", 30)
+        duration_quality = min(duration / target_duration, 1.0) if target_duration > 0 else 1.0
+
+        recommendation = f"Training in Zone '{dominant_zone['name']}' (Effekt: {dominant_zone['effect']})."
+        if duration < target_duration:
+            recommendation += f" Für optimale Wirkung: {target_duration}min empfohlen."
+
+        return {
+            "enabled": True,
+            "dominant_zone": dominant_zone.get("id"),
+            "dominant_zone_name": dominant_zone.get("name"),
+            "avg_hr_percent": round(avg_hr_percent, 1),
+            "zone_color": dominant_zone.get("color"),
+            "zone_effect": dominant_zone.get("effect"),
+            "duration_quality": round(duration_quality, 2),
+            "target_duration_min": target_duration,
+            "actual_duration_min": duration,
+            "recommendation": recommendation
+        }
+
+
+class TrainingEffectsEvaluator:
+    """
+    Evaluates which abilities are trained by the activity.
+    """
+
+    @staticmethod
+    def evaluate(
+        effects_config: Dict,
+        activity: Dict,
+        intensity_zone_result: Optional[Dict] = None
+    ) -> Dict:
+        """
+        Evaluates training effects (abilities trained).
+
+        Args:
+            effects_config: training_effects configuration
+            activity: Activity data
+            intensity_zone_result: Result from intensity zone evaluation
+
+        Returns:
+            {
+                "enabled": bool,
+                "abilities_trained": [...],
+                "total_training_load": float
+            }
+        """
+        if not effects_config.get("enabled", False):
+            return {"enabled": False}
+
+        abilities_trained = []
+
+        # Use default effects if no conditional matching
+        default_effects = effects_config.get("default_effects", {})
+        primary_abilities = default_effects.get("primary_abilities", [])
+        secondary_abilities = default_effects.get("secondary_abilities", [])
+
+        # Calculate quality factor (simplified for now)
+        quality_factor = 1.0
+
+        # Primary abilities
+        for ability in primary_abilities:
+            abilities_trained.append({
+                "category": ability["category"],
+                "ability": ability["ability"],
+                "intensity": ability["intensity"],
+                "quality": quality_factor,
+                "contribution": ability["intensity"] * quality_factor,
+                "type": "primary"
+            })
+
+        # Secondary abilities
+        for ability in secondary_abilities:
+            abilities_trained.append({
+                "category": ability["category"],
+                "ability": ability["ability"],
+                "intensity": ability["intensity"],
+                "quality": quality_factor * 0.7,  # Secondary = 70%
+                "contribution": ability["intensity"] * quality_factor * 0.7,
+                "type": "secondary"
+            })
+
+        total_training_load = sum(a["contribution"] for a in abilities_trained)
+
+        return {
+            "enabled": True,
+            "abilities_trained": abilities_trained,
+            "total_training_load": round(total_training_load, 2),
+            "metabolic_focus": effects_config.get("metabolic_focus", []),
+            "muscle_groups": effects_config.get("muscle_groups", [])
+        }