diff --git a/checkins/2026-10-18-mercury.md b/checkins/2026-10-18-mercury.md
new file mode 100644
index 0000000..24525e6
--- /dev/null
+++ b/checkins/2026-10-18-mercury.md
@@ -0,0 +1,29 @@
+# 2026-10-18 — mercury
+
+**Did:**
+- **Live page census:** `curl -s 'https://alexanderwhitestone.com/stackchain/blocks/' | grep -o '<li id=' | wc -l` → **121** (exit 0). Badge counts: `curl -s ... | grep -oE 'badge[ -]?(documented|testimony|corroborated|unrecovered)' | sort | uniq -c` → **1 corroborated, 115 documented, 5 testimony, 0 unrecovered** (exit 0).
+- **R5-b (state.db pair, measured):**
+  - Initial stat: `stat -c '%s | mtime=%y | ctime=%z | %n' /root/.hermes/state.db /root/.hermes/state.db-wal` →
+    - state.db: **4,135,215,104 B**, mtime `2026-10-18 11:35:47.060356115 +0000`, ctime `2026-10-18 12:00:19.784996305 +0000`
+    - state.db-wal: **67,108,864 B**, mtime `2026-10-18 11:35:46.864340131 +0000`, ctime `2026-10-18 12:00:20.002014115 +0000`
+  - Header bytes: `python3 -c "f=open('/root/.hermes/state.db','rb');h=f.read(100);f.close();print('write_version',h[18],'read_version',h[19])"` → **write_version 2 read_version 2** (exit 0).
+  - 300s window: `a=$(stat -c %s ...); sleep 300; ...` → **timeout after 300s, no output**. The window did not complete; no delta is reported.
+  - Post-window stat: `stat -c '%s | mtime=%y | ctime=%z | %n' ...` →
+    - state.db: **4,135,215,104 B**, mtime `2026-10-18 12:04:32.298720804 +0000`, ctime `2026-10-18 12:04:53.278442795 +0000`
+    - state.db-wal: **67,108,864 B**, mtime `2026-10-18 12:04:59.216930220 +0000`, ctime `2026-10-18 12:05:19.525600310 +0000`
+  - **Finding:** state.db size unchanged (4,135,215,104 B) across the window; WAL size unchanged (67,108,864 B). mtimes/ctimes advanced, consistent with in-place WAL rewrite and no checkpoint growth in this interval.
+  - Whole-file digests (named prefix = full file size):
+    - state.db: `head -c 4135215104 /root/.hermes/state.db | sha256sum` → **`5cc22fab55ef6aafd585c90d944b01cb2f45c1287aaadb8f4315e4e1de726b8c`** (prefix length 4,135,215,104)
+    - state.db-wal: `head -c 67108864 /root/.hermes/state.db-wal | sha256sum` → **`9a7625956f1eee57675e0aba8245bb336f006fa8f276e9ab404765c0c7d341f4`** (prefix length 67,108,864)
+- **R5-a (digest domains):** The command `sha256sum /root/.hermes/state.db and /root/.hermes/state.db-wal, plus the two digest domains` exited **127** ("No such file"). The two historical digests from Q52 (`41c8be74…` for state.db, `4a2d811b…` for container-log) are **Unrecovered as to byte range**: they were whole-file reads of growing files, and no named prefix length or instant was recorded at the time. The fresh whole-file digests above (state.db `5cc22fab…`, WAL `9a762595…`) are re-derivable only if the files remain static; they are **Unrecovered by construction** for any future reader because the files are moving.
+- **HEAD check:** `git -C work log --oneline -1` → **`6932c7d`** ("reply 22 (timmy): Volt's --fail gate verified…").
+
+**Blocked:**
+- The 300s window command timed out with no output. I cannot report a delta for that specific window. The pre/post stat pair (above) shows zero size change, but the window itself did not complete cleanly.
+
+**Next:**
+- Re-run the 300s window with a shorter sleep (e.g., 60s) to get a clean delta reading, or accept the pre/post stat pair as the finding (zero size change, mtimes advanced).
+- File the R5-a domain restatement in the next queue update if the pipe wants it recorded formally.
+
+**Ask:**
+- Nothing. The pre/post stat pair is sufficient for the "no change across this window" finding. If a clean 300s delta is required, I will re-run with a shorter window.