Sandbox MCP

#!/usr/bin/env python3 """ Enhanced Sandbox System Demonstration This script demonstrates the complete enhanced sandbox system with: 1. Enhanced artifact management with categorization 2. Improved REPL with IPython support 3. Complete Manim support with virtual environment integration 4. Performance monitoring and caching 5. Detailed error handling and feedback Usage: python enhanced_sandbox_demo.py """ import sys import asyncio import json from pathlib import Path # Add the sandbox package to path sys.path.insert(0, str(Path(__file__).parent.parent / "src")) from sandbox.sdk import PythonSandbox async def demo_enhanced_artifact_management(): """Demo enhanced artifact management system.""" print("🎨 Enhanced Artifact Management System Demo") print("=" * 50) sandbox_cm = await PythonSandbox.create_local() async with sandbox_cm as sandbox: await sandbox.start() # Create various types of artifacts code = """ import matplotlib.pyplot as plt import numpy as np from PIL import Image import json # Create a plot x = np.linspace(0, 10, 100) y = np.sin(x) plt.figure(figsize=(8, 6)) plt.plot(x, y, 'b-', linewidth=2) plt.title('Sine Wave') plt.xlabel('X') plt.ylabel('Y') plt.grid(True) plt.show() # Create an image img = Image.new('RGB', (200, 200), color='red') img.save('artifacts/images/sample_image.png') # Create data file data = {'x': x.tolist(), 'y': y.tolist()} with open('artifacts/data/sine_data.json', 'w') as f: json.dump(data, f) print("Created various artifacts!") """ result = await sandbox.run(code) output = await result.output() print(f"Execution output: {output}") # Get artifact report report = sandbox.get_artifact_report() print(f"\n📊 Artifact Report:") print(f"Total artifacts: {report['total_artifacts']}") print(f"Total size: {report['total_size'] / 1024:.2f} KB") for category, info in report['categories'].items(): print(f" {category}: {info['count']} files ({info['size'] / 1024:.2f} KB)") # Get categorized artifacts categorized = sandbox.categorize_artifacts() print(f"\n📁 Categorized Artifacts:") for category, files in categorized.items(): if files: print(f" {category}:") for file_info in files: print(f" - {file_info['name']} ({file_info['size']} bytes)") # Get human-readable summary summary = sandbox.get_artifact_summary() print(f"\n📝 Artifact Summary:") print(summary) async def demo_manim_support(): """Demo enhanced Manim support.""" print("\n🎬 Enhanced Manim Support Demo") print("=" * 50) sandbox_cm = await PythonSandbox.create_local() async with sandbox_cm as sandbox: await sandbox.start() # Create a simple Manim animation manim_code = """ from manim import * class SimpleAnimation(Scene): def construct(self): # Create a circle circle = Circle(radius=2, color=BLUE) circle.set_fill(BLUE, opacity=0.5) # Create text text = Text("Hello, Manim!", font_size=48) text.set_color(WHITE) text.move_to(UP * 2.5) # Animate self.play(Create(circle)) self.play(Write(text)) self.play(circle.animate.set_color(RED)) self.play(text.animate.scale(1.5)) self.wait(2) """ result = await sandbox.run(manim_code) output = await result.output() print(f"Manim code execution: {output}") # Get Manim-specific artifacts manim_artifacts = sandbox.get_manim_artifacts() print(f"\n🎥 Manim Artifacts:") for artifact in manim_artifacts: print(f" - {artifact['name']}: {artifact['size']} bytes") async def demo_performance_monitoring(): """Demo performance monitoring and caching.""" print("\n⚡ Performance Monitoring Demo") print("=" * 50) sandbox_cm = await PythonSandbox.create_local() async with sandbox_cm as sandbox: await sandbox.start() # Execute some code multiple times to show caching code = """ import time import numpy as np # Some computational work data = np.random.randn(1000, 1000) result = np.linalg.svd(data) print(f"SVD computation completed") """ # Run first time print("First execution:") result1 = await sandbox.run(code) stats1 = sandbox.get_performance_stats() print(f"Cache hits: {stats1['cache_hits']}, Cache misses: {stats1['cache_misses']}") # Run second time (should use cache) print("\nSecond execution:") result2 = await sandbox.run(code) stats2 = sandbox.get_performance_stats() print(f"Cache hits: {stats2['cache_hits']}, Cache misses: {stats2['cache_misses']}") print(f"Cache hit ratio: {stats2['cache_hit_ratio']:.2%}") # Get execution history history = sandbox.get_execution_history(limit=5) print(f"\n📈 Execution History (last 5):") for i, entry in enumerate(history): print(f" {i+1}. Success: {entry['result']['success']}, " f"Time: {entry['execution_time']:.3f}s") async def demo_error_handling(): """Demo enhanced error handling and feedback.""" print("\n🚨 Enhanced Error Handling Demo") print("=" * 50) sandbox_cm = await PythonSandbox.create_local() async with sandbox_cm as sandbox: await sandbox.start() # Test import error code_with_import_error = """ import nonexistent_module print("This won't run") """ print("Testing import error:") result = await sandbox.run(code_with_import_error) if result.exception: print(f"Error handled: {result.exception}") # Test syntax error code_with_syntax_error = """ def broken_function( print("Missing closing parenthesis") """ print("\nTesting syntax error:") result = await sandbox.run(code_with_syntax_error) if result.exception: print(f"Error handled: {result.exception}") # Test runtime error code_with_runtime_error = """ x = 1 / 0 """ print("\nTesting runtime error:") result = await sandbox.run(code_with_runtime_error) if result.exception: print(f"Error handled: {result.exception}") async def demo_session_management(): """Demo session management and persistence.""" print("\n💾 Session Management Demo") print("=" * 50) sandbox_cm = await PythonSandbox.create_local() async with sandbox_cm as sandbox: await sandbox.start() # Set up some session variables setup_code = """ import numpy as np import matplotlib.pyplot as plt # Create some session variables session_data = { 'user_name': 'Demo User', 'project_name': 'Enhanced Sandbox Demo', 'data': np.random.randn(100) } counter = 0 print(f"Session initialized: {session_data['project_name']}") """ await sandbox.run(setup_code) # Save session sandbox.save_session() print(f"Session saved with ID: {sandbox.session_id}") # Use session variables use_session_code = """ counter += 1 print(f"Counter: {counter}") print(f"User: {session_data['user_name']}") print(f"Data shape: {session_data['data'].shape}") # Create a plot using session data plt.figure(figsize=(10, 6)) plt.hist(session_data['data'], bins=20, alpha=0.7) plt.title(f"{session_data['project_name']} - Data Distribution") plt.xlabel('Value') plt.ylabel('Frequency') plt.show() """ result = await sandbox.run(use_session_code) output = await result.output() print(f"Session usage result: {output}") # Get execution info info = sandbox.get_execution_info() print(f"\n🔍 Execution Environment Info:") print(f"Python version: {info['python_version']}") print(f"Virtual env: {info['virtual_env']}") print(f"Project root: {info['project_root']}") print(f"Artifacts dir: {info['artifacts_dir']}") async def demo_cleanup_management(): """Demo cleanup and management features.""" print("\n🧹 Cleanup Management Demo") print("=" * 50) sandbox_cm = await PythonSandbox.create_local() async with sandbox_cm as sandbox: await sandbox.start() # Create various artifacts code = """ import matplotlib.pyplot as plt import numpy as np from PIL import Image # Create multiple plots for i in range(3): plt.figure(figsize=(6, 4)) plt.plot(np.random.randn(100)) plt.title(f'Plot {i+1}') plt.show() # Create multiple images for i in range(2): img = Image.new('RGB', (100, 100), color=(i*50, 100, 200)) img.save(f'artifacts/images/image_{i}.png') print("Created multiple artifacts") """ await sandbox.run(code) # Show initial artifact count report = sandbox.get_artifact_report() print(f"Initial artifacts: {report['total_artifacts']}") # Cleanup specific type cleaned_plots = sandbox.cleanup_artifacts_by_type('plots') print(f"Cleaned {cleaned_plots} plot artifacts") # Show updated count report = sandbox.get_artifact_report() print(f"Remaining artifacts: {report['total_artifacts']}") # Clear cache sandbox.clear_cache() print("Cache cleared") # Get final stats stats = sandbox.get_performance_stats() print(f"Final stats: {stats}") async def main(): """Run all demos.""" print("🚀 Enhanced Sandbox System - Complete Demo") print("=" * 60) try: await demo_enhanced_artifact_management() await demo_manim_support() await demo_performance_monitoring() await demo_error_handling() await demo_session_management() await demo_cleanup_management() print("\n✅ All demos completed successfully!") print("\nKey Features Demonstrated:") print("✓ Enhanced artifact management with categorization") print("✓ Improved Manim support with virtual environment") print("✓ Performance monitoring and caching") print("✓ Comprehensive error handling") print("✓ Session management and persistence") print("✓ Cleanup and management tools") except Exception as e: print(f"\n❌ Demo failed: {e}") import traceback traceback.print_exc() if __name__ == "__main__": asyncio.run(main())