docs: D1 finish recorded (rationale real-or-absent, reveal once-on-view, 10/80 chip coverage)

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01QJs13VsyiSKYQP6rj3NNmc
2026-07-31 06:18:42 -04:00
parent 085e8a3a63
commit a0501f99c0
2 changed files with 45 additions and 0 deletions
@@ -687,3 +687,22 @@ translateY(-14px) / 6px ring). Audit correction: the `#0E0E14` card token was al

 317 suites / 3,946 tests green, web build exit 0. **Not done, carried in D1:** row-hover rationale,
 IntersectionObserver reveal, team-gradient chips.
+
+---
+
+# D1 FINISH — 2026-07-31 (row anatomy: rationale · reveal · team chips)
+
+Backend untouched. Three pure modules, unit-locked:
+
+- **`rowRationale.js`** — the hover "why" is the grade's OWN `reasoning` + `kill_conditions_triggered`
+  (VERIFIED present on live grades, built from real l5/l20/gap/home/defense/rest), or **NULL**. No
+  generic fallback; a locked/tier-gated reasoning counts as absent rather than paraphrased.
+- **`reveal.js`** — IntersectionObserver, fires **once then unobserves**, reuses D1-A's `bootDelayMs`
+  (one source of truth for the 60ms stagger), and reveals immediately when the API is absent so
+  content is never hidden.
+- **`teamChips.js`** — Rev-3 chip geometry + the 1/.86/.64/.48 ramp. **Colour coverage is 10 of ~80
+  teams** (the artboard's full set, verbatim); every other team renders honest-neutral rather than a
+  guessed colour.
+
+318 suites / 3,961 tests green, web build exit 0. **Mounting the modules into the live row components
+is a follow-up**; the visuals go to the Chrome audit.