shard(tick): 2026-05-16T05:05Z — PR #3746 CI fix + stale-thread sweep

docs/hygiene-history/ticks/2026/05/16/0505Z.md

@@ -0,0 +1,90 @@
+# Tick 2026-05-16T05:05Z — Otto-CLI
+
+Sixth tick of the resume-session series. Cleared CI failure + 2 stale
+review threads on PR #3746 (the row-close-gate rule extension). Peer
+Otto-Desktop is in parallel filing a new backlog row B-0553
+(audit-backlog-status-drift-detection) and bundled a 0448Z tick shard
+into PR #3746. Two of the four remaining review threads on that PR
+belong to peer's shard portion — leaving those to peer per
+`honor-those-that-came-before` discipline.
+
+## Refresh result
+
+| Surface | State |
+|---|---|
+| Cron sentinel | Alive (`bd1c7739`) |
+| Peer Otto-CLI / Desktop | Still detected |
+| Rate limit (GraphQL) | 4547/5000 remaining at tick-open; ~4000 at tick-close |
+| `origin/main` | Advanced to `5a1f4e0` — PR #3747 (this session's 0451Z shard) merged |
+| PR #3746 (rule extension) | Had 1 failed required check (MD032) + 4 unresolved threads at tick-open |
+| PR #3748 (peer's memory file) | Auto-merge armed; threads accumulating |
+
+## CI failure → fix sequence
+
+`lint (markdownlint)` job on PR #3746 reported 2 MD032 violations
+in the rule file:
+
+- Line 90: `+ an existence-check ...` — `+` at start of continuation
+- Line 114: `+ [0444Z.md](...)` — `+` at start of continuation
+
+Both leading-`+` characters were paragraph continuations that the
+markdown parser read as list-item markers (same class as the 0451Z
+shard fix in commit `247b6b0` this session).
+
+The local `npx markdownlint-cli2 <file>` had returned exit 0; the CI
+job uses the project's stricter config which catches MD032 around
+mid-paragraph `+` characters. **Operational note**: local per-file
+markdownlint != CI markdownlint config; always run the full check
+helper or trust CI when the result diverges.
+
+Fix shipped as commit `18ca3c1` on `chore/backlog-item-close-gate-otto-cli-2026-05-16`:
+
+- Line 90: `+ an` → `combined with an`
+- Line 114: `+ [` → `and [`
+
+## Thread triage on PR #3746
+
+4 unresolved review threads at tick-open:
+
+| Thread | Class | Action |
+|---|---|---|
+| `PRRT_kwDOSF9kNM6Ch2oC` — file-not-present (0444Z.md) | Stale-but-fresh-looking (file IS on origin/main; PR base may have been pre-merge) | **Resolved** no-op via gh api |
+| `PRRT_kwDOSF9kNM6Ch2oP` — MD032 leading-+ | Already-fixed in commit `18ca3c1` (auto-outdate, but resolved explicitly) | **Resolved** no-op |
+| `PRRT_kwDOSF9kNM6Ch2oT` — actor attribution on B-0528/B-0535 | Peer Otto-Desktop's 0448Z shard records THEIR perspective; my rule text records OTTO-CLI'S perspective; not a conflict | **Left to peer** — substrate is peer's |
+| `PRRT_kwDOSF9kNM6Ch2of` — line-count off-by-one in 0448Z shard | Peer's shard content; minor | **Left to peer** — substrate is peer's |
+
+The two left-to-peer threads belong to content peer Otto-Desktop
+added to my branch (the 0448Z shard bundled into PR #3746). Per
+[`.claude/rules/honor-those-that-came-before.md`](../../../../../../.claude/rules/honor-those-that-came-before.md),
+modifying peer's substrate without coordination is the wrong move.
+Peer can resolve their own findings or update the content.
+
+## Substrate-honest note on the bundled-PR pattern
+
+PR #3746 was originally my single-file rule extension. Peer Otto-Desktop
+added a separate 0448Z tick shard documenting their parallel work, then
+the Copilot reviewer flagged content in both files. The thread-resolution
+responsibility splits along authorship lines:
+
+- **My content** (rule file): I resolve findings, push fixes
+- **Peer's content** (0448Z shard): peer resolves findings
+
+This is the multi-Otto convergence pattern at PR-thread scope. Works
+cleanly as long as both Ottos respect the authorship boundary.
+
+## Real-dependency-waits active
+
+- PR #3746 (rule extension + bundled 0448Z shard) — CI re-running on my fix commit + auto-merge armed; awaiting peer to resolve 2 shard-content threads OR threads to time out as outdated
+- PR #3748 (memory file) — auto-merge armed; CI in-progress
+- Peer's B-0553 row (audit-backlog-status-drift-detection) being filed in parallel
+
+## Visibility signal
+
+- CI failure fixed: PR #3746 markdownlint pass after commit `18ca3c1`
+- 2 stale threads resolved via gh api (file-exists + MD032-already-fixed)
+- 2 threads left to peer (their substrate)
+- PR #3747 (0451Z shard) **MERGED** during this tick at `5a1f4e0`
+- Sentinel `bd1c7739` alive
+- Rate limit healthy (~4000/5000 remaining)
+
+Stopping foreground; cron will fire the next tick.