docs: Sprint 16.1 — _DAILY_LIMIT 140 → 45 follow-up across all 6 running docs

Sprint 16.1 (commit 11cfd46) is a one-line fix that lowers
_DAILY_LIMIT in backend/app/api/recipe_search.py:48 from
140.0 to 45.0. The 140 value was set assuming Spoonacular's
free tier is 150 pts/day; Sprint 15 round 1 proved the real
cap is 50 pts/day. The gate now triggers at 45 (5pt safety
margin), preventing the user from making requests that
would 503 after a 402 upstream roundtrip.

This commit updates the 6 running docs that track sprints:

- .agent/plan.md — Sprint 16.1 section appended to the
  Sprint 16 sections.
- .agent/context.md — Sprint 16.1 decisions + file:line
  references added.
- Review/sprint16-verification.md — Sprint 16.1 section
  appended (one-line change + verification).
- Review/ui-nielsen-audit.md — Sprint 16.1 paragraph added
  to the Sprint 16 status block.
- fix-ui-audit.md — T9.6 added to the Sprint 16 section.
- Review/handoff-ui-audit.md — TL;DR Sprint 16.1 line
  added, Last-updated footer updated.
- docs/HANDOFF.md — Tracking docs reference updated to
  include Sprint 16.1, Last-updated footer updated.

All 6 docs now reflect Sprint 16.1.

Review/ui-nielsen-audit.md

@@ -159,6 +159,8 @@ The app looks polished on the surface (Tailwind palette, clean cards, working to
 >
 > **Sprint 16 status (in progress 2026-06-08, code complete, awaiting commit + push):** Fix Sprint 13 latent bug — every `/api/llm/plan` call has returned `picked_count=0` since 2026-06-05 because `kimi-k2.6:cloud` is a reasoning model that burns the `max_tokens` budget on internal `reasoning` and never produces the JSON answer. The library fill (Sprint 6+) silently took over every time. Discovered by adding a temp debug log and seeing `raw_response=''` with `finish_reason: length`. Two-line fix: `backend/app/config.py:38` switches `OLLAMA_MODEL` from `kimi-k2.6:cloud` to `gpt-oss:20b` (OpenAI's open-source 20B non-reasoning model); `backend/app/api/llm_plan.py:117` bumps `max_tokens` from 800 to 4000 (21 picks × ~100 chars + reasoning = ~2100+ chars; 4000 gives 2x headroom); `backend/.env` (or `docker-compose` env) also updated so the container reads `gpt-oss:20b`. `frontend/src/api/llm.test.ts` (NEW, 4 cases) — Vitest contract test on the LLM response shape. **Live verification:** 5/5 test weeks return `picked_count` 15-21 (was 0 before). **Tracking:** `Review/sprint16-verification.md`. 11/11 tests pass, `npm run build` green. No new runtime deps. No schema change. No UI change.
 >
+> **Sprint 16.1 (2026-06-08):** one-line `_DAILY_LIMIT` follow-up. `backend/app/api/recipe_search.py:48` — `_DAILY_LIMIT: float = 45.0` (was 140.0). The 140 cap assumed Spoonacular free tier was 150 pts/day; Sprint 15 round 1 proved the real cap is 50. The gate now triggers at 45 (5pt safety margin), preventing the user from making requests that would 503 after a 402 upstream roundtrip. Documented in `Review/sprint16-verification.md` (Sprint 16.1 section).
+>
 >
 > **Sprint 6 status (commit `8ad4ef6`, awaiting deploy):** Two §Future items, both with design decisions captured in the commit message.
 > - **F3** Bulk 'add checked to pantry' on ShoppingList. Backend `POST /api/pantry/bulk` accepts `{items: HomePantryCreate[]}` and returns per-item status (`added` / `updated` / `skipped`) with totals. Per-item failure model: unknown ingredient → `skipped` with reason, not a 4xx. Frontend ShoppingList gains a primary `Add N to pantry` button next to the existing Reset button; toast reports `added X, updated Y, skipped Z`; only the items that actually landed are removed from the checked Set. **Scope decision:** ShoppingList only (the checked Set was the natural substrate; Pantry would need new multi-select UI).