diff --git a/docs/CURRENT-STATE.md b/docs/CURRENT-STATE.md
index 9e4de8e..4afd600 100644
--- a/docs/CURRENT-STATE.md
+++ b/docs/CURRENT-STATE.md
@@ -2998,8 +2998,20 @@
   nine-file figure would omit five.
   **LEDGER MACHINE-DERIVED BLOCK RE-SEEDED** -- it had been carrying the 2026-07-27 seed
   (21 open SEC, next-free D-138) while the per-session summaries below it carried current
-  figures, which is the one drift that block exists to prevent. Now 28 open SEC / D-141 /
-  DOCFIX-208, each added SEC row verified against the register rather than inferred.
+  figures, which is the one drift that block exists to prevent. Now 28 open SEC, D next-free
+  141, DOCFIX next-free 208, each added SEC row verified against the register rather than
+  inferred. **NOTE THE FORM, because getting it wrong is a self-inflicted defect:** these are
+  written as `DOCFIX next-free 208`, NOT as a hyphenated `DOCFIX-NNN` token. `ledger-scan`
+  derives the high-water mark by matching the hyphenated form anywhere in prose, so writing
+  an identifier-shaped token ABOVE the real high-water mark inflates the counter for every
+  later session. This entry did exactly that twice: the first draft wrote the next-free
+  number in hyphenated form and pushed the derived high-water to 208 / next-free 209, and
+  the first CORRECTION then reproduced the same token inside the sentence explaining the
+  mistake. That is the standing numbering rule (hardened in DOCFIX-174) violated one
+  paragraph after being restated, then violated again while being written up. Highest
+  ACTUALLY assigned is 207 (this session's preflight plan-count fix). Caught both times by
+  re-running `ledger-scan` instead of trusting the edit -- the rule is unenforceable by
+  reading, because the defect looks exactly like correct bookkeeping.
 - Project: Omega Cloud, VR1 DC-DC rehearsal -- a two-DC + Office1-headend
   virtual rehearsal on KVM (vcloud host), rehearsing the future bare-metal
   Roosevelt deployment (D-100, `docs/design-decisions.md:1946`).