Add custom bytecode implementation

[schultyy]

This PR introduces a custom bytecode implementation for the Muster VM, improving performance and memory usage.

Key Changes

• Implements a custom bytecode format for all VM instructions
• Adds instruction serialization/deserialization
• Pre-computes jump positions for labels
• Optimizes VM execution by removing unnecessary cloning
• Introduces integer formatting support for printf (%d)
• Adds comprehensive test coverage for bytecode operations

Technical Details

• Each instruction is serialized into a compact byte format
• Label positions are pre-computed and stored in a jump map
• VM execution now operates directly on bytes instead of instruction enums
• Added configuration options for print and remote call queue sizes

Performance Improvements

• Reduced memory usage by removing instruction cloning
• Faster jump operations using pre-computed positions
• More efficient string handling in VM operations

Breaking Changes

• None. All changes are internal to the VM implementation

Test Coverage

Added extensive test cases for all bytecode operations including:

• Push operations (string and int)
• Jump operations
• Printf with integers
• Variable operations
• Stack operations
• Sleep and I/O operations