Igloo MCP

"""Comprehensive integration tests for evolve_report MCP tool. P0 Priority: Critical test coverage for production readiness. """ import uuid from pathlib import Path import pytest from igloo_mcp.config import Config, SnowflakeConfig from igloo_mcp.living_reports.service import ReportService from igloo_mcp.mcp.exceptions import ( MCPSelectorError, MCPValidationError, ) from igloo_mcp.mcp.tools.evolve_report import EvolveReportTool @pytest.mark.asyncio class TestEvolveReportToolIntegration: """Integration tests for evolve_report tool covering critical workflows.""" @pytest.fixture def config(self): """Create test configuration.""" return Config(snowflake=SnowflakeConfig(profile="TEST_PROFILE")) @pytest.fixture def report_service(self, tmp_path: Path): """Create report service with temp storage.""" return ReportService(reports_root=tmp_path / "reports") @pytest.fixture def evolve_tool(self, config, report_service): """Create evolve report tool instance.""" return EvolveReportTool(config, report_service) @pytest.fixture def test_report_id(self, report_service): """Create a test report and return its ID.""" report_id = report_service.create_report( title="Test Report", template="empty", tags=["test"], ) return report_id async def test_evolve_report_success_add_section_with_inline_insights(self, evolve_tool, test_report_id): """Test successful evolution: add section with inline insights.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add revenue analysis section", proposed_changes={ "sections_to_add": [ { "title": "Revenue Analysis", "order": 0, "insights": [ {"summary": "Revenue grew 25% YoY", "importance": 9}, {"summary": "Q4 was strongest quarter", "importance": 8}, ], } ] }, ) assert result["status"] == "success" assert result["summary"]["sections_added"] == 1 assert result["summary"]["insights_added"] == 2 # audit_action_id removed from tool response async def test_evolve_report_success_modify_section_title(self, evolve_tool, test_report_id, report_service): """Test successful evolution: modify existing section title.""" # Setup: Add a section first section_id = None outline = report_service.get_report_outline(test_report_id) if outline.sections: section_id = outline.sections[0].section_id else: # Create a section await evolve_tool.execute( report_selector=test_report_id, instruction="Add test section", proposed_changes={"sections_to_add": [{"title": "Original Title", "order": 0}]}, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Modify section title result = await evolve_tool.execute( report_selector=test_report_id, instruction="Update section title", proposed_changes={"sections_to_modify": [{"section_id": section_id, "title": "Updated Title"}]}, ) assert result["status"] == "success" assert result["summary"]["sections_modified"] == 1 # Verify: Check title was updated updated_outline = report_service.get_report_outline(test_report_id) assert updated_outline.sections[0].title == "Updated Title" async def test_evolve_report_dry_run_validates_without_applying(self, evolve_tool, test_report_id, report_service): """Test dry run mode: validates changes without applying them.""" # Get initial outline initial_outline = report_service.get_report_outline(test_report_id) initial_section_count = len(initial_outline.sections) # Dry run: Attempt to add section result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add test section", proposed_changes={"sections_to_add": [{"title": "Dry Run Section", "order": 0}]}, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result assert "estimated_outline_version" in result["preview"] # Verify: No changes were applied after_outline = report_service.get_report_outline(test_report_id) assert len(after_outline.sections) == initial_section_count async def test_evolve_report_selector_error_invalid_report_id(self, evolve_tool): """Test selector error: non-existent report ID.""" fake_id = f"rpt_{uuid.uuid4()}" with pytest.raises(MCPSelectorError) as exc_info: await evolve_tool.execute( report_selector=fake_id, instruction="Test", proposed_changes={"sections_to_add": [{"title": "Test", "order": 0}]}, ) error = exc_info.value assert error.selector == fake_id assert error.error == "not_found" # hints may not always be in error dict async def test_evolve_report_selector_error_ambiguous_title(self, evolve_tool, report_service): """Test selector error: ambiguous title matches multiple reports.""" # Create two reports with similar titles report_service.create_report(title="Q1 Report", tags=["q1"]) report_service.create_report(title="Q1 Report Analysis", tags=["q1"]) with pytest.raises(MCPSelectorError) as exc_info: await evolve_tool.execute( report_selector="Q1", instruction="Test", proposed_changes={"sections_to_add": [{"title": "Test", "order": 0}]}, ) error = exc_info.value assert error.error in ("ambiguous", "not_found") # Implementation detail # Candidates list depends on selector implementation async def test_evolve_report_validation_error_invalid_response_detail(self, evolve_tool, test_report_id): """Test validation error: invalid response_mode parameter.""" with pytest.raises(MCPValidationError) as exc_info: await evolve_tool.execute( report_selector=test_report_id, instruction="Test", proposed_changes={"sections_to_add": [{"title": "Test", "order": 0}]}, response_mode="invalid", # Should be minimal/standard/full ) error = exc_info.value assert "Invalid response_mode" in error.message assert error.error_code == "VALIDATION_ERROR" assert any("minimal" in hint for hint in error.hints) async def test_evolve_report_validation_error_invalid_proposed_changes_schema(self, evolve_tool, test_report_id): """Test validation error: proposed_changes doesn't match schema.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Test", proposed_changes={ "sections_to_add": [ { "title": "Test Section", "order": "invalid", # Should be int, not str } ] }, ) assert result["status"] == "validation_failed" assert "validation_errors" in result async def test_evolve_report_conflicting_status_change(self, evolve_tool, test_report_id): """Test validation error: conflicting status_change values.""" with pytest.raises(MCPValidationError) as exc_info: await evolve_tool.execute( report_selector=test_report_id, instruction="Test", proposed_changes={"status_change": "archived"}, status_change="active", # Conflicting value ) error = exc_info.value assert "Conflicting status_change" in error.message async def test_evolve_report_response_detail_minimal(self, evolve_tool, test_report_id): """Test response_detail='minimal' returns compact response.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={"sections_to_add": [{"title": "Test Section", "order": 0}]}, response_detail="minimal", ) assert result["status"] == "success" # Minimal response should have summary but minimal details assert "summary" in result assert result["summary"]["sections_added"] == 1 # Should NOT have verbose section details in minimal mode assert "report_id" in result # audit_action_id removed from tool response async def test_evolve_report_response_detail_full(self, evolve_tool, test_report_id): """Test response_detail='full' returns comprehensive response.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={"sections_to_add": [{"title": "Test Section", "order": 0}]}, response_detail="full", ) assert result["status"] == "success" assert "summary" in result assert "report_id" in result # Full response includes changes_applied details assert "changes_applied" in result async def test_evolve_report_batch_operations(self, evolve_tool, test_report_id): """Test batch operations: add multiple sections and insights together.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add comprehensive analysis", proposed_changes={ "sections_to_add": [ { "title": "Executive Summary", "order": 0, "insights": [{"summary": "Strong growth in Q4", "importance": 9}], }, { "title": "Detailed Analysis", "order": 1, "insights": [ {"summary": "Revenue up 30%", "importance": 8}, {"summary": "Costs down 10%", "importance": 7}, ], }, ] }, ) assert result["status"] == "success" assert result["summary"]["sections_added"] == 2 assert result["summary"]["insights_added"] == 3 async def test_evolve_report_status_change(self, evolve_tool, test_report_id, report_service): """Test status change: archive report via evolve_report.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Archive report", proposed_changes={"status_change": "archived"}, ) assert result["status"] == "success" # Verify: Check report status was updated outline = report_service.get_report_outline(test_report_id) assert outline.metadata.get("status") == "archived" async def test_evolve_report_remove_section(self, evolve_tool, test_report_id, report_service): """Test section removal.""" # Setup: Add a section first await evolve_tool.execute( report_selector=test_report_id, instruction="Add test section", proposed_changes={"sections_to_add": [{"title": "Section to Remove", "order": 0}]}, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Remove the section result = await evolve_tool.execute( report_selector=test_report_id, instruction="Remove section", proposed_changes={"sections_to_remove": [section_id]}, ) assert result["status"] == "success" # Verify: Section was removed updated_outline = report_service.get_report_outline(test_report_id) assert len(updated_outline.sections) == 0 async def test_evolve_report_title_change(self, evolve_tool, test_report_id, report_service): """Test report title change.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Update report title", proposed_changes={"title_change": "Updated Report Title"}, ) assert result["status"] == "success" # Verify: Title was updated outline = report_service.get_report_outline(test_report_id) assert outline.title == "Updated Report Title" async def test_evolve_report_metadata_updates(self, evolve_tool, test_report_id, report_service): """Test metadata updates via proposed_changes.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Update metadata", proposed_changes={ "metadata_updates": { "owner": "data_team", "priority": "high", "custom_field": "custom_value", } }, ) assert result["status"] == "success" # Verify: Metadata was updated outline = report_service.get_report_outline(test_report_id) assert outline.metadata.get("owner") == "data_team" assert outline.metadata.get("priority") == "high" assert outline.metadata.get("custom_field") == "custom_value" async def test_evolve_report_empty_proposed_changes_fallback(self, evolve_tool, test_report_id): """Test empty proposed_changes triggers fallback generation.""" # When proposed_changes is empty, tool should generate changes from instruction # This is a fallback path for simple use cases result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add revenue section", proposed_changes={}, # Empty - should trigger fallback ) # The fallback should still work (may generate minimal/no-op changes) assert result["status"] in ("success", "validation_failed") async def test_evolve_report_request_id_tracking(self, evolve_tool, test_report_id): """Test request_id is tracked through the operation.""" custom_request_id = str(uuid.uuid4()) result = await evolve_tool.execute( report_selector=test_report_id, instruction="Test request tracking", proposed_changes={"sections_to_add": [{"title": "Test Section", "order": 0}]}, request_id=custom_request_id, ) assert result["status"] == "success" # Request ID should be in the response context or traceable in logs async def test_evolve_report_audit_trail(self, evolve_tool, test_report_id, report_service): """Test that evolution creates audit trail.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section for audit test", proposed_changes={"sections_to_add": [{"title": "Audit Test Section", "order": 0}]}, ) assert result["status"] == "success" # audit_action_id removed from tool response # Verify: Audit event was created storage = report_service.global_storage.get_report_storage(test_report_id) audit_events = storage.load_audit_events() assert len(audit_events) > 0 # Find the evolution event evolution_events = [e for e in audit_events if e.action_type == "evolve"] assert len(evolution_events) > 0 # ========================================================================= # content_merge_mode Integration Tests (Issue #101) # ========================================================================= async def test_evolve_section_content_merge_mode_replace(self, evolve_tool, test_report_id, report_service): """Test content_merge_mode='replace' replaces content entirely.""" # Setup: Add a section with initial content await evolve_tool.execute( report_selector=test_report_id, instruction="Add section with content", proposed_changes={ "sections_to_add": [ { "title": "Test Section", "order": 0, "content": "Original paragraph one.\n\nOriginal paragraph two.", } ] }, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Replace content entirely result = await evolve_tool.execute( report_selector=test_report_id, instruction="Replace section content", proposed_changes={ "sections_to_modify": [ { "section_id": section_id, "content": "Completely new content.", "content_merge_mode": "replace", } ] }, ) assert result["status"] == "success" # Verify: Content is completely replaced outline = report_service.get_report_outline(test_report_id) assert outline.sections[0].content == "Completely new content." assert "Original" not in outline.sections[0].content async def test_evolve_section_content_merge_mode_append(self, evolve_tool, test_report_id, report_service): """Test content_merge_mode='append' adds content at end.""" # Setup: Add a section with initial content await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={ "sections_to_add": [ { "title": "Append Test", "order": 0, "content": "First paragraph.", } ] }, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Append content result = await evolve_tool.execute( report_selector=test_report_id, instruction="Append to section", proposed_changes={ "sections_to_modify": [ { "section_id": section_id, "content": "Second paragraph added.", "content_merge_mode": "append", } ] }, ) assert result["status"] == "success" # Verify: New content is appended outline = report_service.get_report_outline(test_report_id) content = outline.sections[0].content assert "First paragraph." in content assert "Second paragraph added." in content # First should come before second assert content.index("First paragraph.") < content.index("Second paragraph added.") async def test_evolve_section_content_merge_mode_prepend(self, evolve_tool, test_report_id, report_service): """Test content_merge_mode='prepend' adds content at beginning.""" # Setup: Add a section with initial content await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={ "sections_to_add": [ { "title": "Prepend Test", "order": 0, "content": "Original content.", } ] }, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Prepend content result = await evolve_tool.execute( report_selector=test_report_id, instruction="Prepend to section", proposed_changes={ "sections_to_modify": [ { "section_id": section_id, "content": "New introduction.", "content_merge_mode": "prepend", } ] }, ) assert result["status"] == "success" # Verify: New content is prepended outline = report_service.get_report_outline(test_report_id) content = outline.sections[0].content assert "Original content." in content assert "New introduction." in content # New intro should come before original assert content.index("New introduction.") < content.index("Original content.") async def test_evolve_section_content_merge_mode_merge_with_placeholders( self, evolve_tool, test_report_id, report_service ): """Test content_merge_mode='merge' with placeholder-based merging.""" # Setup: Add a section with structured content initial_content = ( "# Introduction\n\nOpening paragraph.\n\n# Analysis\n\nAnalysis content.\n\n# Conclusion\n\nFinal thoughts." ) await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={ "sections_to_add": [ { "title": "Merge Test", "order": 0, "content": initial_content, } ] }, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Merge with placeholders - insert new content while keeping existing new_content = "// ... keep above ...\n\n# New Analysis\n\nNew analysis paragraph.\n\n// ... keep below ..." result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add to analysis section", proposed_changes={ "sections_to_modify": [ { "section_id": section_id, "content": new_content, "content_merge_mode": "merge", } ] }, ) assert result["status"] == "success" # Verify: Original content should still be there with new content merged outline = report_service.get_report_outline(test_report_id) content = outline.sections[0].content # Check that original sections are preserved assert "# Introduction" in content or "Opening paragraph." in content assert "# New Analysis" in content or "New analysis paragraph." in content async def test_evolve_section_content_default_merge_mode_is_replace( self, evolve_tool, test_report_id, report_service ): """Test that default content_merge_mode is 'replace'.""" # Setup: Add a section with initial content await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={ "sections_to_add": [ { "title": "Default Mode Test", "order": 0, "content": "Original content.", } ] }, ) outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Test: Modify without specifying content_merge_mode (should default to replace) result = await evolve_tool.execute( report_selector=test_report_id, instruction="Update section", proposed_changes={ "sections_to_modify": [ { "section_id": section_id, "content": "New content only.", # No content_merge_mode specified - should default to "replace" } ] }, ) assert result["status"] == "success" # Verify: Content is replaced (original not present) outline = report_service.get_report_outline(test_report_id) assert outline.sections[0].content == "New content only." assert "Original" not in outline.sections[0].content @pytest.mark.asyncio class TestEvolveReportToolEdgeCases: """Edge case tests for evolve_report tool.""" @pytest.fixture def config(self): return Config(snowflake=SnowflakeConfig(profile="TEST_PROFILE")) @pytest.fixture def report_service(self, tmp_path: Path): return ReportService(reports_root=tmp_path / "reports") @pytest.fixture def evolve_tool(self, config, report_service): return EvolveReportTool(config, report_service) @pytest.fixture def test_report_id(self, report_service): return report_service.create_report(title="Edge Case Test", tags=["test"]) async def test_evolve_report_large_batch_operations(self, evolve_tool, test_report_id): """Test handling of large batch operations (stress test).""" # Create 20 sections with 5 insights each = 100 total insights sections_to_add = [] for i in range(20): sections_to_add.append( { "title": f"Section {i}", "order": i, "insights": [{"summary": f"Insight {j} for section {i}", "importance": 5} for j in range(5)], } ) result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add comprehensive analysis", proposed_changes={"sections_to_add": sections_to_add}, ) assert result["status"] == "success" assert result["summary"]["sections_added"] == 20 assert result["summary"]["insights_added"] == 100 async def test_evolve_report_unicode_content(self, evolve_tool, test_report_id): """Test handling of Unicode and special characters.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add international content", proposed_changes={ "sections_to_add": [ { "title": "国际化测试 🌍", "order": 0, "insights": [ { "summary": "Revenue in 日本 grew 25% 📈", "importance": 8, } ], } ] }, ) assert result["status"] == "success" async def test_evolve_report_modify_nonexistent_section(self, evolve_tool, test_report_id): """Test error when modifying non-existent section.""" fake_section_id = str(uuid.uuid4()) result = await evolve_tool.execute( report_selector=test_report_id, instruction="Modify non-existent section", proposed_changes={"sections_to_modify": [{"section_id": fake_section_id, "title": "New Title"}]}, ) assert result["status"] == "validation_failed" assert "validation_errors" in result # ===== PHASE 4: Dry-Run Preview Tests (#135) ===== @pytest.mark.asyncio class TestEvolveReportDryRunPreview: """Tests for enhanced dry-run preview with rendered content and diffs - Issue #135.""" @pytest.fixture def config(self): """Create test configuration.""" return Config(snowflake=SnowflakeConfig(profile="TEST_PROFILE")) @pytest.fixture def report_service(self, tmp_path: Path): """Create report service with temp storage.""" return ReportService(reports_root=tmp_path / "reports") @pytest.fixture def evolve_tool(self, config, report_service): """Create evolve report tool instance.""" return EvolveReportTool(config, report_service) @pytest.fixture def test_report_id(self, report_service): """Create a test report and return its ID.""" return report_service.create_report( title="Dry Run Test Report", template="empty", tags=["test"], ) async def test_dry_run_returns_preview_for_new_sections(self, evolve_tool, test_report_id): """Dry run should include rendered preview for new sections.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add new section", proposed_changes={ "sections_to_add": [ { "title": "New Analysis Section", "order": 0, "content": "This is the section content with **bold** text.", } ] }, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result preview = result["preview"] assert preview["sections_to_add"] == 1 # Check for rendered preview of new sections if "rendered_preview" in preview: assert "new_sections" in preview["rendered_preview"] new_sections = preview["rendered_preview"]["new_sections"] assert len(new_sections) == 1 assert new_sections[0]["title"] == "New Analysis Section" # Preview should contain markdown assert "preview_markdown" in new_sections[0] async def test_dry_run_returns_preview_for_new_insights(self, evolve_tool, test_report_id): """Dry run should include rendered preview for new insights.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add new insight", proposed_changes={ "insights_to_add": [ { "summary": "Revenue grew 25% year-over-year driven by enterprise sales", "importance": 9, } ] }, constraints={"skip_citation_validation": True}, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result preview = result["preview"] assert preview["insights_to_add"] == 1 # Check for rendered preview of new insights if "rendered_preview" in preview: assert "new_insights" in preview["rendered_preview"] new_insights = preview["rendered_preview"]["new_insights"] assert len(new_insights) == 1 assert new_insights[0]["importance"] == 9 # Preview should contain formatted insight assert "preview" in new_insights[0] assert "Insight:" in new_insights[0]["preview"] async def test_dry_run_returns_modifications_diff(self, evolve_tool, test_report_id, report_service): """Dry run should include before/after diff for modifications.""" # First add a section to modify await evolve_tool.execute( report_selector=test_report_id, instruction="Add section to modify later", proposed_changes={ "sections_to_add": [ { "title": "Original Title", "order": 0, "content": "Original content here.", } ] }, ) # Get the section ID outline = report_service.get_report_outline(test_report_id) section_id = outline.sections[0].section_id # Now do a dry run to modify it result = await evolve_tool.execute( report_selector=test_report_id, instruction="Modify section title", proposed_changes={ "sections_to_modify": [ { "section_id": section_id, "title": "Updated Title", } ] }, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result preview = result["preview"] assert preview["sections_to_modify"] == 1 # Check for modifications with before/after if "modifications" in preview: mods = preview["modifications"] assert len(mods) >= 1 # Find the title modification title_mod = next((m for m in mods if m.get("field") == "title"), None) if title_mod: assert title_mod["before"] == "Original Title" assert title_mod["after"] == "Updated Title" assert title_mod["type"] == "section" async def test_dry_run_returns_insight_modification_diff(self, evolve_tool, test_report_id, report_service): """Dry run should include before/after diff for insight modifications.""" # First add an insight to modify await evolve_tool.execute( report_selector=test_report_id, instruction="Add insight to modify later", proposed_changes={ "insights_to_add": [ { "summary": "Original insight summary", "importance": 5, } ] }, constraints={"skip_citation_validation": True}, ) # Get the insight ID outline = report_service.get_report_outline(test_report_id) insight_id = outline.insights[0].insight_id # Now do a dry run to modify it result = await evolve_tool.execute( report_selector=test_report_id, instruction="Modify insight importance", proposed_changes={ "insights_to_modify": [ { "insight_id": insight_id, "importance": 9, } ] }, constraints={"skip_citation_validation": True}, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result preview = result["preview"] assert preview["insights_to_modify"] == 1 # Check for modifications with before/after if "modifications" in preview: mods = preview["modifications"] importance_mod = next((m for m in mods if m.get("field") == "importance"), None) if importance_mod: assert importance_mod["before"] == 5 assert importance_mod["after"] == 9 assert importance_mod["type"] == "insight" async def test_dry_run_preview_truncates_long_content(self, evolve_tool, test_report_id): """Dry run preview should truncate long content for token efficiency.""" long_content = "A" * 1000 # 1000 characters result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section with long content", proposed_changes={ "sections_to_add": [ { "title": "Long Content Section", "order": 0, "content": long_content, } ] }, dry_run=True, ) assert result["status"] == "dry_run_success" # If rendered_preview exists, check truncation if "rendered_preview" in result.get("preview", {}): new_sections = result["preview"]["rendered_preview"].get("new_sections", []) if new_sections: preview_md = new_sections[0].get("preview_markdown", "") # Should be truncated to reasonable length (500 chars in implementation) assert len(preview_md) <= 600 # Some buffer for markdown formatting async def test_dry_run_estimated_version(self, evolve_tool, test_report_id): """Dry run should include estimated new version number.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section", proposed_changes={"sections_to_add": [{"title": "Test Section", "order": 0}]}, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result assert "estimated_outline_version" in result["preview"] # Should be current version + 1 assert result["preview"]["estimated_outline_version"] >= 1 async def test_dry_run_no_changes_applied(self, evolve_tool, test_report_id, report_service): """Dry run should not actually apply any changes.""" # Get original state original_outline = report_service.get_report_outline(test_report_id) original_section_count = len(original_outline.sections) # Do dry run result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section (dry run)", proposed_changes={"sections_to_add": [{"title": "Should Not Appear", "order": 0}]}, dry_run=True, ) assert result["status"] == "dry_run_success" # Verify no changes were made after_outline = report_service.get_report_outline(test_report_id) assert len(after_outline.sections) == original_section_count async def test_dry_run_combined_operations_preview(self, evolve_tool, test_report_id): """Dry run should show preview for multiple operation types.""" result = await evolve_tool.execute( report_selector=test_report_id, instruction="Add section and insight", proposed_changes={ "sections_to_add": [{"title": "New Section", "order": 0}], "insights_to_add": [{"summary": "New insight", "importance": 7}], }, constraints={"skip_citation_validation": True}, dry_run=True, ) assert result["status"] == "dry_run_success" assert "preview" in result preview = result["preview"] assert preview["sections_to_add"] == 1 assert preview["insights_to_add"] == 1