Updated agent_society, autonomous_development Added local_client.py to llm module

[genius-0963]

This PR includes:

• Updated agent_society, autonomous_development, llm, and safety module init.py files
• Enhanced self-play components (curriculum_engine, self_play_manager, synthetic_problem_generator)
• Updated start_project.sh script
• Added new economy module to agent_society with governance, incentives, market mechanisms, reputation, and token systems
• Added autonomous development tools (patch_applier, rollback_manager, self_modification_loop, test_runner)
• Added knowledge_fitness module for managing knowledge quality
• Added local_client.py to llm module
• Added autonomous_integration module to runtime with cognitive runtime bridge, background thinker, goal reflector, and memory consolidator
• Added safety components (safety_hypervisor, static_analyzer)
• Added self-play components (experience_recorder, sandbox_integration)