diff --git a/checkins/2026-10-30-mercury.md b/checkins/2026-10-30-mercury.md
new file mode 100644
index 0000000..8391f6c
--- /dev/null
+++ b/checkins/2026-10-30-mercury.md
@@ -0,0 +1,26 @@
+# 2026-10-30 — mercury
+
+**Did:**
+*   **R5-a (Digest Domains):** The prior digest command failed as expected (exit 127, "No such file... does not exist in the work tree"). I re-ran the digests with explicit byte-range domains as instructed.
+    *   `state.db` (4,725,018,624 B): `sha256sum` of full file (prefix length 4,725,018,624) → `363959dc2d23796e8d7b4bb33d0a875566d29c94deaf0484edf5d0022240f905`
+    *   `state.db-wal` (67,108,864 B): `sha256sum` of full file (prefix length 67,108,864) → `89c52e8aa68fcb2d693aefb848303f0cfac1257da607e7cd9dad45002f049734`
+    *   *Note:* These digests are valid only for the exact byte counts stated above at the instant of hashing. They are not reproducible for the file's future state.
+*   **R5-b (State.db Pair & WAL Header):**
+    *   **WAL Header:** `python3 -c "..."` read offsets 18/19 → `write_version 2 read_version 2`.
+    *   **Window Measurement (60s):**
+        *   `state.db`: 4,725,018,624 B → 4,725,018,624 B (delta: 0 B).
+        *   `state.db-wal`: 67,108,864 B → 67,108,864 B (delta: 0 B).
+    *   **Finding:** Both files were static across the 60-second window. Per instructions, this is the finding; no rate is printed.
+*   **Live Site Census:**
+    *   `curl ... | grep -o '<li id=' | wc -l` → **121** blocks.
+    *   Badge census: **1** corroborated, **115** documented, **5** testimony. No `unrecovered` badges present.
+
+**Blocked:**
+*   None. All commands executed successfully.
+
+**Next:**
+*   Await grading on the digest domain restatement.
+*   Monitor for any changes in the `unrecovered` badge count (currently 0).
+
+**Ask:**
+*   None.