/teach

name: teach
description: "Explain a subsystem to a learner by composing how and why — the walkthrough cited to code, the decisions that shaped it, paced with a prediction check. For onboarding and knowledge transfer."
aliases: []
metadata:
  credits:
    compose-how-why: pstack
    author: Lauren Tan
    url: "https://github.com/cursor/plugins/tree/main/pstack"
    port: pstack-claude
    port-author: Michael Denyer
    port-url: "https://github.com/michael-denyer/pstack-claude"

Teach

Teaching is `how` and `why` composed for a learner: what the subsystem does and the reasons it is shaped that way, in an order a person can absorb.

Workflow

1. **Ask the learner's baseline**: one question — familiar with the domain, the language, or neither. Size the explanation to the answer. 2. **How first** (`how` skill): the walkthrough, entry to exit, cited to file:line, at the depth the baseline supports. 3. **Then why** (`why` skill): the 2-4 decisions that shaped it, each with its evidence and the alternative it beat. Human rulings quoted verbatim. 4. **Check the landing**: ask the learner to predict one behavior — "if X happens, what does the subsystem do?" A correct prediction ends the lesson; a wrong one names the layer to redo. The check is for pacing, not grading.

Rules

• One new concept per paragraph; jargon defined at first use in one clause.
• Cite code for behavior, cite records for reasons; label anything you

inferred.

• Stop when the learner can predict behavior — completeness is not the goal,

transfer is.