feat(pedagogy): saddle-extrema audit + capability additions (§11.7–8)

[bradleypmartin]

Sub-issue of #233.

Per-scene pedagogical coverage audit for saddle-extrema against APPM 2350 §11.7–8, and the resulting capability additions. Sub-epic — audit-phase findings get broken out into real sub-sub-issues at plan-doc time. Starter audit content TBD — to be filled in during the per-scene plan-doc phase.

Pedagogical anchor

§11.7–8 framing: second-derivative test, critical-point classification. Does the scene make the classification logic — Hessian eigenstructure, the discriminant D = f_xx·f_yy − f_xy² — visible and felt, not just declared?

Note the scene already received substantive pedagogical-content moves in v0.8/v0.9: critical-point markers at analytic CPs (#179/#186), local-quadratic-approximation overlay (#180/#187), CP-aware snap detents (#200/#212). The audit should account for what's already there before identifying gaps.

Audit (defer to plan doc)

• Walk the scene with the pedagogical-anchor question in hand; produce the starter list of gaps.
• Compare current capability against §11.7–8 syllabus content; flag missing or thin coverage given what's already shipped.
• Identify candidate capability additions; triage into v1.0-scope vs. v1.x-deferred.

Composition

Paired with the saddle-extrema narrative (#231). Plan together.