Design a comprehensive meta-prompt framework for building an AI-driven task delegation system. This system should intelligently assign tasks to individuals or teams based on their skills, availability, and task priorities.  The meta-prompt should encompass several sub-prompts to handle various aspects of the system, including task intake, skill assessment, availability tracking, task assignment logic, and performance monitoring.  Consider different task types (e.g., simple, complex, urgent), individual skill levels (beginner, intermediate, expert), and potential constraints (e.g., deadlines, dependencies).  The output should be a structured document outlining the complete system architecture, including the individual prompts and their interconnections, data flow, and potential integration with existing project management tools.  The system should be designed for scalability and adaptability to changing team structures and task priorities.  Include examples of specific prompts for each stage of the task delegation process, such as a prompt for assessing individual skills or a prompt for determining task urgency. The final output should provide a clear evaluation metric for measuring the efficiency and effectiveness of the AI-driven task delegation system.  Consider scenarios where tasks need to be re-assigned due to unforeseen circumstances or delays.  The system should be robust enough to handle such situations gracefully.