Framework for effective AI collaboration defining task decomposition, context management, abstraction selection, and automation philosophy.

• Divide complex tasks into small, independently verifiable steps; AI performs significantly better with clear, bounded instructions than large ambiguous requests
• Keep context fresh and single-purpose; use separate conversations for different topics and apply HANDOFF.md summaries when conversations grow long to prevent context drift
• Adjust abstraction level by situation: use high-level "vibe coding" for prototypes and rapid iteration, deep-dive line-by-line review for production code, bugs, and security work
• Default to plan mode (70-90% of work) for analysis and review before execution; use execute mode only for simple, well-validated tasks in safe environments
• Automate any task repeated 3+ times; verify all outputs through tests, code review, and self-verification before deployment