issues: flag #336/#222/#321 as merged-but-unresolved, #322 as in-progress

docs/outstanding-issues-inbox/1505ba13-1b2a-47c9-b6cc-03b060f7a878.json

@@ -0,0 +1,11 @@
+{
+  "version": 2,
+  "id": "1505ba13-1b2a-47c9-b6cc-03b060f7a878",
+  "createdOn": "2026-08-18",
+  "action": "update",
+  "payload": {
+    "id": "#222",
+    "detail": "PR #2072 merged into main 2026-08-18 citing this row, but only shipped a cosmetic font-weight change (font-extrabold to font-semibold) in mode-home-template.tsx and a SearchResultsEmptyState class-alignment tweak in search-results-header-band.tsx — not the PageHeader/EmptyState convergence decision this row actually asks for. Stays open: the in-scope-or-not decision still hasn't been made.",
+    "baseRowFingerprint": "ba6621ae785bf636071d361b64b416ca35d8e6bc893eb6bb1ccef662dae46465"
+  }
+}

docs/outstanding-issues-inbox/779bc0e9-c28f-40ab-abda-cfdcf93ea752.json

@@ -0,0 +1,11 @@
+{
+  "version": 2,
+  "id": "779bc0e9-c28f-40ab-abda-cfdcf93ea752",
+  "createdOn": "2026-08-18",
+  "action": "update",
+  "payload": {
+    "id": "#321",
+    "detail": "PR #2072 merged into main 2026-08-18 citing this row (differentials compare page / filmstrip / DocumentViewer control alignment), but its changed-files list never touched any of those files. Entirely unaddressed by this merge despite being cited in the PR title. Stays open.",
+    "baseRowFingerprint": "b1b6db157f9d95ca18ab31ba8ccc709dc28cf603838d599b19f1b694f88b4ac5"
+  }
+}

docs/outstanding-issues-inbox/8e28c783-d655-4480-852c-bafaf0b3c07b.json

@@ -0,0 +1,11 @@
+{
+  "version": 2,
+  "id": "8e28c783-d655-4480-852c-bafaf0b3c07b",
+  "createdOn": "2026-08-18",
+  "action": "update",
+  "payload": {
+    "id": "#322",
+    "detail": "PR #2069 (gemini/clinical-medication-graph-dedup) is open and targets this row. As of 2026-08-18 it reconciles both Warfarin catalogue records to carry the same interaction row set (the row's core ask), but its own new coverage test (tests/medication-interaction-lexicon-coverage.test.ts) still fails: the two duplicate records don't cross-resolve each other by name. That's a content/authoring decision (cross-link vs. merge-to-one-canonical-record) left for clinical/authoring review, not yet fixed. Stays open pending that PR landing correctly — flagging so a future session doesn't open a duplicate PR for the same dedup (see #292).",
+    "baseRowFingerprint": "3ff5c10503f0942a5b38f3f7e3a770308493c08ee3e64d686a9668978f0b033a"
+  }
+}

docs/outstanding-issues-inbox/a39af37c-b4dd-4e4b-84dc-395d86937615.json

@@ -0,0 +1,11 @@
+{
+  "version": 2,
+  "id": "a39af37c-b4dd-4e4b-84dc-395d86937615",
+  "createdOn": "2026-08-18",
+  "action": "update",
+  "payload": {
+    "id": "#336",
+    "detail": "PR #2072 merged into main 2026-08-18 citing this row, and the --breakpoint-phone/-tablet/-desktop plus --bp-phone/-tablet/-desktop tokens now exist in src/app/globals.css (confirmed on main) — but the row's required decision (record raw-vs-named-tokens choice in docs/design-system/GATES.md) was never made. GATES.md has not been touched since PR #1990 and contains no mention of breakpoints or #336. The row's own stop rule ('do not migrate one window ahead of the decision') was not honored; the merge shipped the migration without the decision record. Stays open: still needs the GATES.md decision entry.",
+    "baseRowFingerprint": "757b5240cff62bd5071d12e4f2ada7e7c3faebb8018843b45a8675913b7442d9"
+  }
+}