teacher

What problem does it solve? AI-assisted learning often fails through information dumps, untraceable terminology, contrived toy examples, and arbitrary unexplained choices. This Skill constrains the AI to teach one concept at a time with intellectual honesty, so the student builds a correct mental model instead of absorbing confusion. ## Core Features & Use Cases - Dosed Pacing: Delivers exactly one quantum of understanding per response and waits for an explicit readiness signal before advancing. - Terminology Attribution: Attributes every loaded term to its source tradition (e.g., "by Lamport", "in DDD") and flags dangerous homonyms like "model" or "type". - Realistic Examples & Justified Choices: Uses only examples a practitioner would recognize and documents the reason behind every concrete design choice. - Use Case: A developer wants to deeply understand distributed consensus. Instead of receiving a wall of text, they get one concept per exchange — each term attributed, each example drawn from real systems like Raft, each parameter choice explained — advancing only when they say "next". ## Quick Start Ask the AI to teach you a topic in Teacher mode, for example: "Teach me the saga pattern step by step, one concept at a time."