diff --git a/docs/outstanding-issues-inbox/51b7e3d4-9c38-4fd7-9aa8-70d46086fde9.json b/docs/outstanding-issues-inbox/51b7e3d4-9c38-4fd7-9aa8-70d46086fde9.json
new file mode 100644
index 0000000000..2ce689f7b2
--- /dev/null
+++ b/docs/outstanding-issues-inbox/51b7e3d4-9c38-4fd7-9aa8-70d46086fde9.json
@@ -0,0 +1,12 @@
+{
+  "version": 2,
+  "id": "51b7e3d4-9c38-4fd7-9aa8-70d46086fde9",
+  "createdOn": "2026-08-21",
+  "action": "update",
+  "payload": {
+    "id": "#VTEW3W",
+    "summary": "therapyBtn still dresses 13 raw controls across 7 therapy files with no shared equivalent",
+    "detail": "CORRECTION 2026-08-21, superseding the update applied earlier the same day: the count of 21 across 8 files recorded by that update was WRONG, and the 13 across 7 files measured by the request cancelled as c6a9756d is RIGHT. The discrepancy flagged in that cancellation is now resolved by direct re-measurement on main rather than by choosing between the two numbers. Cause of the error: the earlier update counted raw textual occurrences of therapyBtn, which includes one import statement per consuming file plus the export in controls.ts, and then compared that total against a baseline expressed in CALL SITES. That is an apples-to-oranges comparison and it overstated the remaining work. Measured per file, excluding the import line in each and excluding the definition: brief-screen 1, compare-screen 1, detail-screen 1, pathways-screen 1, recommend-screen 1, sheets-screen 7, therapy-card 1 = 13 call sites across 7 .tsx files, with the export at controls.ts:26. Against the 30 call sites recorded 2026-08-18 at adf93a75, that is a real reduction of 17, not the 9 the bad arithmetic implied. The design question and the stop rule are unchanged: nothing shared covers list rows, disclosure headers, section toggles and chips, and this must not be resolved by converting the remaining sites to Button variants.",
+    "baseRowFingerprint": "731faf770b9414607177a3e24ae530d98479a5c5f65547da063825563bf711f4"
+  }
+}
diff --git a/docs/outstanding-issues-inbox/5d987a3d-3869-4864-8af0-de0a6145b682.json b/docs/outstanding-issues-inbox/5d987a3d-3869-4864-8af0-de0a6145b682.json
new file mode 100644
index 0000000000..8b981cf409
--- /dev/null
+++ b/docs/outstanding-issues-inbox/5d987a3d-3869-4864-8af0-de0a6145b682.json
@@ -0,0 +1,11 @@
+{
+  "version": 2,
+  "id": "5d987a3d-3869-4864-8af0-de0a6145b682",
+  "createdOn": "2026-08-21",
+  "action": "done",
+  "payload": {
+    "id": "#SZGPAH",
+    "outcome": "Resolved 2026-08-21. Re-measured first-hand on 2026-08-21 at HEAD e4df772e3712620a8d64dc28843048e32bde88ee: node scripts/run-playwright.mjs --project=chromium-mockups tests/ui-tools-search-mode-mockup.spec.ts reported '16 passed (1.5m)' with exit code 0. Both previously-failing tests are green by name: test 1, 'desktop uses universal search and keeps results beside the selected-tool panel' (spec line 24, 10.9s), and test 9, 'phone filter sheet follows the shared local-filter behavior' (spec line 194, 4.5s). This independently reproduces the run at a341832af cited by the request cancelled as 3eb84c6a-97fa-4b1c-9167-190ba928c200, whose author explicitly recommended that a fresh done request be queued so the evidence would not be lost; that recommendation is what this request carries out. The rows are therefore closed on a reproduced measurement, not on a third-party claim. The advisory lane is no longer red for every UI PR from this spec. Note for the record: the 2026-08-21 update on this row said 'NOT VERIFIED: whether the spec now passes ... Re-measure before closing' - that re-measurement has now been done and is the basis for this closure.",
+    "baseRowFingerprint": "7b4cfefa46485dc95291481287de889362d2241b6bdc18599b969ef11f7ee5b7"
+  }
+}
diff --git a/docs/outstanding-issues-inbox/c1f99542-526a-4218-b8e6-87bf136d7567.json b/docs/outstanding-issues-inbox/c1f99542-526a-4218-b8e6-87bf136d7567.json
new file mode 100644
index 0000000000..b2e5304520
--- /dev/null
+++ b/docs/outstanding-issues-inbox/c1f99542-526a-4218-b8e6-87bf136d7567.json
@@ -0,0 +1,11 @@
+{
+  "version": 2,
+  "id": "c1f99542-526a-4218-b8e6-87bf136d7567",
+  "createdOn": "2026-08-21",
+  "action": "done",
+  "payload": {
+    "id": "#4TBHS8",
+    "outcome": "Resolved 2026-08-21. Re-measured first-hand on 2026-08-21 at HEAD e4df772e3712620a8d64dc28843048e32bde88ee: node scripts/run-playwright.mjs --project=chromium-mockups tests/ui-tools-search-mode-mockup.spec.ts reported '16 passed (1.5m)' with exit code 0. Both previously-failing tests are green by name: test 1, 'desktop uses universal search and keeps results beside the selected-tool panel' (spec line 24, 10.9s), and test 9, 'phone filter sheet follows the shared local-filter behavior' (spec line 194, 4.5s). This independently reproduces the run at a341832af cited by the request cancelled as 1a175782-d9cf-4de6-a67b-d3b2da632c84, whose author explicitly recommended that a fresh done request be queued so the evidence would not be lost; that recommendation is what this request carries out. The rows are therefore closed on a reproduced measurement, not on a third-party claim. The specific failure recorded here - waiting for the exact text '2 showing' inside the tools-search-filter-sheet testid - is gone with the rewritten assertion, and the rewritten test passes. Tracked with #SZGPAH, which named the same spec file and closed on the same run.",
+    "baseRowFingerprint": "3769f42e3e4a417bbb9b3e2c30200a08eaa98eb879a95bc569b3d37dda2e95ca"
+  }
+}