Explains a change or subsystem so a person actually understands it, rather than summarizing it. It runs pstack-how and pstack-why in parallel and weaves their findings into one plain account — what the thing is, how it works, why it is that way — pitched at the person’s pace. The goal is explicitly that they understand it, not that you change anything.
Two mechanisms make it more than a prettier explainer, and both are unusual enough to be worth recording.
Build the picture up diagram by diagram. For anything with three or more moving parts the skill forbids one diagram containing all of them. It requires a short series instead, where each diagram redraws the last and adds a single part, so the reader watches the system assemble. To teach a flow from A to B to C, draw it three times: A to B, then redraw and add C, then redraw and add the return edge. A single all-at-once diagram, especially one saved for the end, is ruled reference, not teaching. The medium is matched to the idea — mermaid where labels carry the meaning, a generated marker-on-whiteboard image where the idea is spatial (layout, overlap, scroll position, a before and after).
Anti-pedagogy-theater rules. No quizzes, no “pause and say it back”, no announcing that a part is important or tricky, no printed framing labels (“the key insight”, “at its core”, “TL;DR”). Give the smallest complete answer first, a sentence or two, then stop and let them respond. Layers get added when asked.
Its cross-framework counterpart is ce-explain — both turn work into a human-facing explainer that feeds understanding forward rather than shipping anything, which is why both sit in stage-learn. The difference is the audience model: ce-explain produces a dense visual artifact, /teach produces a paced conversation.
See Also
- pstack-how · pstack-why — the two skills it runs and blends; it must not redo their work.
- pstack-unslop — every response is written through it.
- ce-explain — the cross-framework counterpart.
- stage-learn — the canonical stage this implements.