slide-plan

slide-plan — per-slide section blocks (Skill 2)

Sub-skill of module-overview-page (Skill 1). When a module has multiple slide decks, each slide reads best as its own section: what the slide covers + the assignments/labs based on it, with that slide embedded right there.

Model

from slide_plan import slide_sections
sections = slide_sections([
  {"title": "📽️ Slide 1 — While Loops",
   "summary_html": "<p>Covers the while-loop pattern: condition, update, sentinel stop.</p>",
   "tasks": [("assignment", 1314443, "Assignment 5-0 — Power Number"),
             ("assignment", 1314445, "Assignment 5-1 — Input Validation")],
   "deck_id": "<SLIDE_1_DECK_ID>"},
  {"title": "📽️ Slide 2 — For Loops", "summary_html": "<p>…</p>",
   "tasks": [("assignment", 1314447, "Assignment 5-3 — Count Random")],
   "deck_id": "<SLIDE_2_DECK_ID>"},
])
# then hand `sections` to Skill 1:
#   make_page(course_slug, slug, overview_html, sections, deck_id=None, ...)

Each dict → a (header, summary_html, tasks, deck_id) section. tasks items = (kind, ref, label) (kind ∈ assignment/page/quiz/url/text) — the SAME item shape Skill 1 renders.

What is AUTHORED vs CANON

• Authored (the AI): each slide's title, summary_html (read the real deck to summarize it — concrete, NO <code>), and which tasks derive from it.
• Canon: the packing (slide_plan.py) + the rendering/embedding (Skill 1's module_overview.py).

How it fits

Skill 1 (module-overview-page) drives the page; for a multi-slide module it calls this to produce the sections. A single-slide module doesn't need this — Skill 1 uses a page-level deck_id + plain category sections instead.