devforgeai 1.0.5 → 1.0.6

package/src/claude/scripts/devforgeai_cli/tests/feedback/test_integration.py

@@ -0,0 +1,402 @@
+"""
+Integration tests for retrospective feedback system
+
+Tests the complete workflow end-to-end with real file I/O,
+YAML persistence, and actual workflow scenarios.
+"""
+
+import pytest
+import tempfile
+import shutil
+from pathlib import Path
+from datetime import datetime, timezone
+
+from devforgeai_cli.feedback.retrospective import (
+    trigger_retrospective,
+    capture_feedback,
+    save_in_progress_state,
+    resume_feedback,
+    detect_rapid_sequence,
+)
+from devforgeai_cli.feedback.skip_tracking import (
+    increment_skip,
+    get_skip_count,
+    reset_skip_count,
+    check_skip_threshold,
+)
+from devforgeai_cli.feedback.aggregation import (
+    aggregate_feedback_by_story,
+    aggregate_feedback_by_epic,
+    detect_patterns,
+    generate_insights,
+    export_quarterly_insights,
+)
+from devforgeai_cli.feedback.longitudinal import (
+    correlate_feedback_across_stories,
+    identify_improvement_trajectories,
+    export_personal_journal,
+)
+
+
+class TestFullWorkflowIntegration:
+    """Test complete workflow from trigger → capture → aggregation"""
+
+    @pytest.fixture
+    def temp_dirs(self):
+        """Create temporary directories for feedback and config"""
+        temp_dir = tempfile.mkdtemp()
+        feedback_dir = Path(temp_dir) / 'feedback'
+        config_dir = Path(temp_dir) / 'config'
+        feedback_dir.mkdir(parents=True)
+        config_dir.mkdir(parents=True)
+
+        yield feedback_dir, config_dir
+
+        shutil.rmtree(temp_dir)
+
+    def test_successful_dev_workflow_end_to_end(self, temp_dirs):
+        """
+        GIVEN user completes /dev STORY-001 successfully
+        WHEN retrospective triggered, feedback captured, and aggregated
+        THEN complete workflow executes without errors
+        """
+        feedback_dir, config_dir = temp_dirs
+
+        # Step 1: Trigger retrospective
+        operation_result = {
+            'status': 'success',
+            'story_id': 'STORY-001',
+        }
+        questions = trigger_retrospective('dev', operation_result)
+
+        assert len(questions) >= 4
+        assert all(hasattr(q, 'question_id') for q in questions)
+
+        # Step 2: Simulate user responses
+        responses = [
+            {'question_id': questions[0].question_id, 'response': 5, 'skip': False},
+            {'question_id': questions[1].question_id, 'response': 'Green', 'skip': False},
+            {'question_id': questions[2].question_id, 'response': 'The workflow was very clear and well-structured.', 'skip': False},
+            {'question_id': questions[3].question_id, 'response': 4, 'skip': False},
+        ]
+
+        # Step 3: Capture feedback
+        result = capture_feedback(
+            responses=responses,
+            story_id='STORY-001',
+            workflow_type='dev',
+            success_status='success',
+            feedback_dir=feedback_dir,
+            epic_id='EPIC-001',
+        )
+
+        assert result['status'] == 'recorded'
+        assert 'feedback_id' in result
+        assert 'file_path' in result
+
+        # Verify file was written
+        assert Path(result['file_path']).exists()
+
+        # Step 4: Aggregate feedback
+        aggregated = aggregate_feedback_by_story(feedback_dir)
+        assert 'STORY-001' in aggregated
+        assert len(aggregated['STORY-001']) == 1
+
+    def test_failed_qa_workflow_with_skip_tracking(self, temp_dirs):
+        """
+        GIVEN user runs /qa that fails 3 times and skips feedback each time
+        WHEN skip threshold reached
+        THEN system detects pattern and suggests improvement
+        """
+        feedback_dir, config_dir = temp_dirs
+        user_id = 'test_user'
+
+        # Simulate 3 failed QA operations with feedback skipped
+        for i in range(3):
+            # Trigger retrospective
+            operation_result = {
+                'status': 'failed',
+                'story_id': f'STORY-00{i+1}',
+                'failure_reason': 'Coverage below threshold',
+            }
+            questions = trigger_retrospective('qa', operation_result)
+
+            # User skips all questions
+            responses = [
+                {'question_id': q.question_id, 'response': '', 'skip': True}
+                for q in questions
+            ]
+
+            # Capture feedback (should be skipped)
+            result = capture_feedback(
+                responses=responses,
+                story_id=f'STORY-00{i+1}',
+                workflow_type='qa',
+                success_status='failed',
+                feedback_dir=feedback_dir,
+                allow_skip=True,
+            )
+
+            assert result['status'] == 'skipped'
+
+            # Increment skip count
+            count = increment_skip(user_id, config_dir)
+            assert count == i + 1
+
+        # Verify threshold detected
+        assert check_skip_threshold(user_id, threshold=3, config_dir=config_dir)
+        assert get_skip_count(user_id, config_dir) == 3
+
+    def test_network_loss_recovery(self, temp_dirs):
+        """
+        GIVEN user starts feedback but connection lost mid-way
+        WHEN they reconnect and resume
+        THEN partial responses are preserved and recovered
+        """
+        feedback_dir, config_dir = temp_dirs
+
+        # Step 1: Start feedback session
+        partial_responses = [
+            {'question_id': 'dev_success_01', 'response': 4, 'skip': False},
+            {'question_id': 'dev_success_02', 'response': 'Red', 'skip': False},
+        ]
+
+        # Save in-progress state
+        state_file = save_in_progress_state(
+            story_id='STORY-001',
+            workflow_type='dev',
+            responses=partial_responses,
+            feedback_dir=feedback_dir,
+        )
+
+        assert state_file.exists()
+
+        # Step 2: Simulate recovery
+        recovered_state = resume_feedback(story_id='STORY-001', feedback_dir=feedback_dir)
+
+        assert recovered_state is not None
+        assert recovered_state['story_id'] == 'STORY-001'
+        assert recovered_state['workflow_type'] == 'dev'
+        assert len(recovered_state['responses']) == 2
+        assert recovered_state['responses'][0]['response'] == 4
+
+    def test_rapid_sequence_detection(self, temp_dirs):
+        """
+        GIVEN user runs /dev STORY-001, then /dev STORY-002 within 30 seconds
+        WHEN checking for rapid sequence
+        THEN system detects rapid sequence
+        """
+        feedback_dir, config_dir = temp_dirs
+
+        # First feedback
+        operation_result_1 = {
+            'status': 'success',
+            'story_id': 'STORY-001',
+        }
+        questions_1 = trigger_retrospective('dev', operation_result_1)
+        responses_1 = [
+            {'question_id': q.question_id, 'response': 'Test', 'skip': False}
+            for q in questions_1
+        ]
+
+        result_1 = capture_feedback(
+            responses=responses_1,
+            story_id='STORY-001',
+            workflow_type='dev',
+            success_status='success',
+            feedback_dir=feedback_dir,
+        )
+
+        # Get timestamp from first feedback
+        import json
+        with open(result_1['file_path'], 'r') as f:
+            first_feedback = json.load(f)
+
+        first_timestamp = datetime.fromisoformat(first_feedback['timestamp'].replace('Z', '+00:00'))
+
+        # Check rapid sequence (should be detected if within 30 seconds)
+        is_rapid = detect_rapid_sequence(first_timestamp, threshold_seconds=30)
+
+        # Since we just created it, current time - first_timestamp should be < 30 seconds
+        assert is_rapid is True
+
+    def test_aggregation_and_insights_generation(self, temp_dirs):
+        """
+        GIVEN multiple users provide feedback across multiple stories
+        WHEN generating insights
+        THEN patterns are detected and actionable recommendations generated
+        """
+        feedback_dir, config_dir = temp_dirs
+
+        # Create 5 feedback sessions with common issue
+        for i in range(5):
+            operation_result = {
+                'status': 'success',
+                'story_id': f'STORY-00{i+1}',
+            }
+            questions = trigger_retrospective('dev', operation_result)
+
+            responses = [
+                {'question_id': questions[0].question_id, 'response': 3, 'skip': False},
+                {'question_id': questions[1].question_id, 'response': 'Red', 'skip': False},
+                {'question_id': questions[2].question_id, 'response': 'tests are hard to write', 'skip': False},
+                {'question_id': questions[3].question_id, 'response': 3, 'skip': False},
+            ]
+
+            capture_feedback(
+                responses=responses,
+                story_id=f'STORY-00{i+1}',
+                workflow_type='dev',
+                success_status='success',
+                feedback_dir=feedback_dir,
+                epic_id='EPIC-001',
+            )
+
+        # Generate insights
+        insights = generate_insights(feedback_dir)
+
+        assert 'recommendations' in insights
+        # With 5 feedback sessions all saying "tests are hard to write",
+        # pattern detection should find this (100% occurrence)
+        recommendations = insights['recommendations']
+        assert len(recommendations) > 0
+
+        # Export quarterly insights
+        output_path = export_quarterly_insights(feedback_dir)
+        assert output_path.exists()
+        assert output_path.name == 'quarterly-insights.md'
+
+    def test_longitudinal_tracking_across_stories(self, temp_dirs):
+        """
+        GIVEN user completes multiple stories over time with improving confidence
+        WHEN analyzing trajectory
+        THEN system detects improvement trend
+        """
+        feedback_dir, config_dir = temp_dirs
+        user_id = 'test_user'
+
+        # Create 5 feedback sessions with increasing confidence
+        for i in range(5):
+            operation_result = {
+                'status': 'success',
+                'story_id': f'STORY-00{i+1}',
+            }
+            questions = trigger_retrospective('dev', operation_result)
+
+            # Confidence increases: 2, 3, 3, 4, 5
+            confidence = min(2 + (i // 2), 5)
+
+            responses = [
+                {'question_id': questions[0].question_id, 'response': confidence, 'skip': False},
+                {'question_id': questions[1].question_id, 'response': 'Red', 'skip': False},
+                {'question_id': questions[2].question_id, 'response': 'Getting better!', 'skip': False},
+                {'question_id': questions[3].question_id, 'response': confidence, 'skip': False},
+            ]
+
+            capture_feedback(
+                responses=responses,
+                story_id=f'STORY-00{i+1}',
+                workflow_type='dev',
+                success_status='success',
+                feedback_dir=feedback_dir,
+                epic_id='EPIC-001',
+            )
+
+        # Correlate feedback across stories
+        correlation = correlate_feedback_across_stories(feedback_dir, user_id)
+        assert 'timeline' in correlation
+        assert len(correlation['timeline']) == 5
+
+        # Identify improvement trajectory
+        trajectories = identify_improvement_trajectories(feedback_dir, user_id)
+        assert 'metrics' in trajectories
+
+        if len(trajectories['metrics']) > 0:
+            metric = trajectories['metrics'][0]
+            assert metric['name'] == 'TDD confidence'
+            # Should be improving (2 → 5)
+            assert metric['start_value'] <= metric['end_value']
+            assert metric['trend'] == 'improving'
+
+        # Export personal journal
+        journal_path = export_personal_journal(feedback_dir, user_id)
+        assert journal_path.exists()
+        assert journal_path.name == 'journal.md'
+
+
+class TestErrorHandlingIntegration:
+    """Test error handling in integrated scenarios"""
+
+    @pytest.fixture
+    def temp_dirs(self):
+        """Create temporary directories"""
+        temp_dir = tempfile.mkdtemp()
+        feedback_dir = Path(temp_dir) / 'feedback'
+        config_dir = Path(temp_dir) / 'config'
+        feedback_dir.mkdir(parents=True)
+        config_dir.mkdir(parents=True)
+
+        yield feedback_dir, config_dir
+
+        shutil.rmtree(temp_dir)
+
+    def test_invalid_story_id_rejected(self, temp_dirs):
+        """
+        GIVEN user provides invalid story ID format
+        WHEN capturing feedback
+        THEN validation fails gracefully
+        """
+        from devforgeai_cli.feedback.validation import validate_story_id
+
+        # Invalid formats
+        assert validate_story_id('story-001') is False  # lowercase
+        assert validate_story_id('STORY001') is False    # no hyphen
+        assert validate_story_id('EPIC-001') is False    # wrong prefix
+
+        # Valid format
+        assert validate_story_id('STORY-001') is True
+
+    def test_insufficient_responses_rejected(self, temp_dirs):
+        """
+        GIVEN user provides only 1 substantive response out of 5 questions
+        WHEN capturing feedback
+        THEN validation error raised
+        """
+        feedback_dir, config_dir = temp_dirs
+
+        responses = [
+            {'question_id': 'q1', 'response': 'Good', 'skip': False},  # Only 1 substantive
+            {'question_id': 'q2', 'response': '', 'skip': True},
+            {'question_id': 'q3', 'response': '', 'skip': True},
+            {'question_id': 'q4', 'response': '', 'skip': True},
+            {'question_id': 'q5', 'response': '', 'skip': True},
+        ]
+
+        with pytest.raises(ValueError, match='At least 2 of 5 questions'):
+            capture_feedback(
+                responses=responses,
+                story_id='STORY-001',
+                workflow_type='dev',
+                success_status='success',
+                feedback_dir=feedback_dir,
+            )
+
+    def test_sensitive_content_detection_integration(self, temp_dirs):
+        """
+        GIVEN user accidentally includes API key in feedback
+        WHEN capturing feedback
+        THEN system detects and flags sensitive content
+        """
+        from devforgeai_cli.feedback.validation import check_sensitive_content
+
+        feedback_with_key = "The API key sk-1234567890abcdefghijklmnopqrst was exposed in logs"
+        is_sensitive, types = check_sensitive_content(feedback_with_key)
+
+        assert is_sensitive is True
+        assert 'secret' in types
+
+        feedback_normal = "The workflow was great and easy to follow"
+        is_sensitive, types = check_sensitive_content(feedback_normal)
+
+        assert is_sensitive is False
+        assert len(types) == 0

package/src/claude/scripts/devforgeai_cli/tests/feedback/test_models.py

@@ -0,0 +1,105 @@
+"""Unit tests for models.py data models.
+
+Tests Question and FeedbackSession dataclasses.
+"""
+
+import pytest
+from devforgeai_cli.feedback.models import Question, FeedbackSession
+
+
+class TestQuestion:
+    """Test Question dataclass."""
+
+    def test_question_creation_valid(self):
+        """Test creating a valid question."""
+        q = Question(
+            question_id='q1',
+            question_text='How confident are you?',
+            response_type='rating',
+            scale='1-5'
+        )
+        assert q.question_id == 'q1'
+        assert q.scale == '1-5'
+
+    def test_rating_question_without_scale_raises_error(self):
+        """Test that rating question without scale raises ValueError (line 24)."""
+        with pytest.raises(ValueError, match="must have scale"):
+            Question(
+                question_id='q1',
+                question_text='Rate this',
+                response_type='rating'
+                # Missing scale - should raise error
+            )
+
+    def test_multiple_choice_without_options_raises_error(self):
+        """Test that multiple choice without options raises ValueError (line 27)."""
+        with pytest.raises(ValueError, match="must have at least 2 options"):
+            Question(
+                question_id='q2',
+                question_text='Choose one',
+                response_type='multiple_choice'
+                # Missing options - should raise error
+            )
+
+    def test_multiple_choice_with_one_option_raises_error(self):
+        """Test that multiple choice with only 1 option raises ValueError (line 27)."""
+        with pytest.raises(ValueError, match="must have at least 2 options"):
+            Question(
+                question_id='q3',
+                question_text='Choose one',
+                response_type='multiple_choice',
+                options=['Only one option']  # Need at least 2
+            )
+
+
+class TestFeedbackSession:
+    """Test FeedbackSession dataclass."""
+
+    def test_feedback_session_creation(self):
+        """Test creating a FeedbackSession."""
+        session = FeedbackSession(
+            feedback_id='fb-1',
+            timestamp='2025-01-01T00:00:00Z',
+            story_id='STORY-001',
+            epic_id='EPIC-001',
+            workflow_type='dev',
+            success_status='success'
+        )
+        assert session.feedback_id == 'fb-1'
+        assert session.story_id == 'STORY-001'
+
+    def test_feedback_session_to_dict(self):
+        """Test converting FeedbackSession to dict."""
+        session = FeedbackSession(
+            feedback_id='fb-1',
+            timestamp='2025-01-01T00:00:00Z',
+            story_id='STORY-001'
+        )
+        data = session.to_dict()
+
+        assert data['feedback_id'] == 'fb-1'
+        assert data['story_id'] == 'STORY-001'
+        assert 'timestamp' in data
+
+    def test_feedback_session_from_dict(self):
+        """Test creating FeedbackSession from dict (line 58)."""
+        data = {
+            'feedback_id': 'fb-2',
+            'timestamp': '2025-01-01T12:00:00Z',
+            'story_id': 'STORY-002',
+            'epic_id': 'EPIC-001',
+            'workflow_type': 'qa',
+            'success_status': 'failed',
+            'questions': [{'question_id': 'q1', 'response': 3}],
+            'metadata': {'duration': 120}
+        }
+
+        session = FeedbackSession.from_dict(data)
+
+        assert session.feedback_id == 'fb-2'
+        assert session.story_id == 'STORY-002'
+        assert session.epic_id == 'EPIC-001'
+        assert session.workflow_type == 'qa'
+        assert session.success_status == 'failed'
+        assert len(session.questions) == 1
+        assert session.metadata['duration'] == 120