diff --git a/docs/hygiene-history/ticks/2026/05/14/2204Z.md b/docs/hygiene-history/ticks/2026/05/14/2204Z.md
new file mode 100644
index 0000000000..1b5388e424
--- /dev/null
+++ b/docs/hygiene-history/ticks/2026/05/14/2204Z.md
@@ -0,0 +1,95 @@
+# Tick 2026-05-14T22:04Z — PR #3276 double-backtick rendering fix per Copilot review
+
+## Refresh (step 1)
+
+- Cron sentinel `12fb713e` live.
+- **#3277** (shard 2202Z minimal-continuation) MERGED as `dcd8475`.
+- **#3276** (shard 2158Z) — wait-ci with 1 new Copilot thread.
+
+## Holding discipline (step 2)
+
+Per `.claude/rules/blocked-green-ci-investigate-threads.md`: investigate.
+
+## Work (step 3) — Copilot rendering catch on broken code span
+
+Copilot catch on 2158Z.md line 28:
+
+> P1: Inline-code formatting looks broken here: the string includes a
+> single backtick after `.claude/` but no matching opening/closing
+> backtick pair, so Markdown may render the remainder of the line as
+> a code span.
+
+The broken line was:
+
+```text
+- Prose describing the bug ("`](.claude/` patterns")
+```
+
+The single-backtick code span `` `](.claude/` `` rendered unpredictably
+because `](` can confuse Markdown parsers expecting link-closing syntax.
+
+Fix: wrap example pattern in double-backticks per the inline-code-span
+discipline codified in tick 2154Z:
+
+> Content containing backticks | Double backticks: `` ``content with
+> `inner` backticks`` ``
+
+Commit `a3dd133` on #3276 branch; thread resolved with reply.
+
+## Verify (step 4)
+
+- `markdownlint-cli2` clean on the fixed 2158Z shard
+- Thread resolved via GraphQL
+- Composite branch-guard + `gh pr create --head` used
+
+## Shard (step 5)
+
+This file.
+
+## CronList (step 6)
+
+Sentinel `12fb713e` armed; one entry, recurring.
+
+## Visibility (step 7)
+
+- **PR #3276** (audit-deferral shard) — fix `a3dd133` pushed, thread
+  resolved, CI re-running; autoMerge stays armed
+- **PR #3277** merged this batch
+- Session running tally: **34 merged + 1 wait-ci** (then plus this shard's PR)
+
+## Aaron signaled off-duty (observed via parallel-Otto substrate)
+
+During this tick's branch-contamination recovery, the contaminating
+branch (`shard/2026-05-14-shadow-session-close-off-duty-aaron-signaled`)
+contained commit `0a9a2e2` by Aaron:
+
+```text
+docs(research/shadow): session-close off-duty signaled by Aaron 2026-05-14
+"off-duty until next signal (shadow*)"
+```
+
+This is an **authorization-source signal** per
+`.claude/rules/mechanical-authorization-check.md` (Aaron is the only
+authorization source). Future ticks should respect this signal:
+
+- **Land in-flight work** (this tick's audit fix + shard)
+- **Not start new substantive PRs** beyond what's already in motion
+- **Continue per-tick discipline** (cron still fires every minute;
+  shards still required) but stay minimal
+- **Wait for the "next signal"** Aaron mentioned
+
+## Notes for future-Otto
+
+**Inline-code-span discipline applied iteratively**: this is the
+SECOND time this session Copilot caught a broken inline code span
+caused by `](...)` content inside single backticks (first was #3271
+line 44 → #3274; this is 2158Z line 28 → fix a3dd133).
+
+Both fixes use double-backticks per the 2154Z discipline table.
+The pattern is reliable: **anytime an inline code span needs to
+contain `](`, use double backticks**.
+
+The auditing mechanization gap is real — and **it's the next
+session's work**, not this session's. Aaron's off-duty signal makes
+that explicit: stop accumulating new work; close out cleanly; wait
+for next signal.