소개
Context Fields is an advanced cognitive framework for Claude Code that improves model performance by applying negative constraints—or 'inhibitions'—rather than just positive instructions. By automatically detecting user intent—whether it's debugging, planning, or creative brainstorming—this skill enforces specific mental models that prevent common LLM pitfalls like hallucinations, over-engineering, or premature execution. It allows Claude to shift into specialized regimes such as adversarial testing, empathetic support, or rigorous technical critique, ensuring the response aligns perfectly with the cognitive demands of the task.
