AI Principles Framework

Produced by Masonsoft AI, the AI Principles Framework for Personal Calibration is a guided prompt kit for professionals who want more from an AI assistant than stylistic consistency. Where a preferences list tells an AI how to format a response, a principles document tells it how you think. This framework builds the latter through structured conversation. The document contains an opening system prompt that, when pasted into a new Claude session, instructs the AI to guide the user through seven sequential layers. The first three address epistemology, professional identity, and relationship to conventional wisdom, drawing out where the user agrees with received opinion in their field and where they dissent. Layer four captures editorial instincts beyond formatting preferences. Layer five identifies genuine limits: the content, tones, and framings the user will not produce regardless of commercial pressure. Layers six and seven address how the user handles intellectual challenge and what they observe that others in their field consistently miss. The result is a personal principles document that can be saved and reused in future AI sessions. The framework includes guidance on reviewing the first draft, correcting inaccuracies, and updating individual layers as professional circumstances change.