docs(PENDING): close PENDING-180 and -184; PENDING-182 ADDENDUM 2; PENDING-146 ADDENDUM 1

PENDING-180 is closed as DISCHARGED by REVIEWED-138, a ruling whose heading
names PENDING-168. That is the second CLASS E mirror in two days, recorded as
such rather than as routine, and recorded at PENDING-146 where the class lives,
with a census of the form text can see (one true instance) and a statement that
the first instance's form is invisible to it by construction.

PENDING-184 is closed on 37a2c86: both control arms shown to fail by mutation,
the live path shown still to record, the live log unchanged under a real run.

PENDING-182 ADDENDUM 2 states that the field survives the collapse of its
stated rationale (PENDING-150 §6b cited for its diagnosis; its outcome went the
other way, to 4d2ae87) and makes PENDING-184 a precondition of the field.

PENDING-146 ADDENDUM 1 also names, without repairing, bare-headed addenda that
sit after PENDING-183 and are attributed to it by every id-keyed reader.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01PJM5fwqp456LDGzqiZXsgu

PENDING.md

@@ -2094,6 +2094,24 @@ PENDING-146 proposes. **No `Awaiting:` line.**
 **Files affected:** authoring convention (`/wrap-up` §1.6 or `~/CLAUDE.md` §Steward-Jurist Interface — the latter is `[ESCALATE]`, steward's hand); `~/dotfiles/scripts/wake-digest.py` for detection.
 **Awaiting:** Steward. ⚠ **Independently of any ruling: Move 1 and Move 2 await direction and have been unable to say so since 2026-08-13; the parent, ADDENDUM 1 and ADDENDUM 2 have been unable to say so since 2026-08-10.**
 
+### PENDING-146 — ADDENDUM 1: CLASS E, mirrored — two instances in two days, recorded where the class lives
+**Date:** 2026-09-11
+
+**Neither mirror instance was recorded here, so the class's own item did not show its recurrence.** The first was classified inside PENDING-139's addendum (*"Filed as PENDING-146's shape rather than as a third item"*), the second inside PENDING-180. The mirror: **a ruling's unit is the id its heading names, while the unit it discharges sits under an id that heading never names.** Same defect as this item, other direction.
+
+| # | found | unit discharged | ruling, and the id its heading names | where the ruling names the unit | recorded at |
+|---|---|---|---|---|---|
+| 1 | 2026-09-10 | PENDING-139 option (a) | REVIEWED-132 — PENDING-173 | nowhere | PENDING-139, second addendum of 2026-09-10 |
+| 2 | 2026-09-11 | PENDING-180, whole | REVIEWED-138 — PENDING-168 | its body: *"PENDING-180 is DISCHARGED."* | PENDING-180 ADDENDUM 2 |
+
+**Same direction both times: discharged on the record, open on the list.** Both were found by a party reading for another purpose — answering a jurist note on 09-10, reading a ruling against the digest's open list on 09-11. **No instrument is aimed at either form.**
+
+**Census for the form text can see** (PENDING-180 ADDENDUM 2): one true instance in `REVIEWED.md` — No. 2 — one false positive, both controls held. **No. 1's form is invisible to any text search over rulings by construction**, since the ruling never names the item. **Two is a floor, not a count.**
+
+⚠ **A placement defect of this item's own family, found while writing this and NOT repaired here.** PENDING-139's two addenda of 2026-09-10 — including the one that classifies instance 1 — carry bare `### ADDENDUM` headings and sit physically after PENDING-183. Every id-keyed reader therefore attributes them to PENDING-183. Two further bare addenda in the same stretch do not state their item in their headings either, and which item owns them is not established here. **The unit problem, one level down: an addendum's item is decided by where it sits, not by what it says.** Moving blocks is relocation under the lossless-relocation gate, not the executor's to do unasked. Named so it can be aimed at.
+
+**Awaiting:** unchanged on the parent.
+
 ---
 
 ## PENDING-147 — The ladder trial's counter is a 30-day ROLLING WINDOW, so `transcripts 84` can never fire and the trial's own evidence is being deleted
@@ -4345,6 +4363,24 @@ chk(f"a real session reads as WRAPPED end-to-end  [{_v.count('wrapped')} of {len
 
 **Awaiting:** Steward authorization on the parent. ⚠ **What remains for a ruling is smaller than when this item was filed:** (c), (b) and (a) are executed and the repair is mutation-verified, so the item is **dischargeable on a ruling that settles its count unit**. The new gap above is a separate decidable question and is **not** bundled here — filing it inside this addendum would reproduce the CLASS E defect PENDING-146 names, in the item that just declined to commit it.
 
+### PENDING-180 — ADDENDUM 2: CLOSED — discharged by REVIEWED-138, whose heading names PENDING-168: the second CLASS E mirror in two days
+**Date:** 2026-09-11
+
+**Discharged by REVIEWED-138** (placed `7c096a2`), which rules: *"**PENDING-180 is DISCHARGED.** Its (c) census → (b) `source_has()` → (a) repair were executed in that order and the repair is mutation-verified."* That ruling's heading names **PENDING-168**, not this item. The open list resolves closure by the id a ruling's heading names, so this item read open for a day with its discharge already placed — the wake digest listed it among 55 open items on 2026-09-11.
+
+**This is PENDING-146's CLASS E, mirrored — the second instance in two days, same defect, same direction.** CLASS E is *"the open list's unit is the ID; the decidable unit is the BLOCK."* The mirror: **the ruling's unit is the id its heading names, while the unit it discharges sits under an id the heading never names.**
+- **First, classified 2026-09-10** in PENDING-139's second addendum of that date: option (a) shipped under REVIEWED-132, whose heading names PENDING-173 and which never mentions PENDING-139 at all.
+- **Second, this item:** the ruling names the discharged item in its body, not its heading.
+- **Same direction both times:** discharged on the record, open on the list. The open list overstated what was open; nothing was understated.
+
+⚠ **Recorded as a second instance, not as routine.** The first being named yesterday made this one no cheaper to notice: it surfaced when REVIEWED-138 was read for another purpose, against the digest's open list. No instrument is aimed at it.
+
+**Census, 2026-09-11 — for the form a text search can see; a floor for the class.** Every REVIEWED entry whose body puts DISCHARGED, discharged, CLOSED, WITHDRAWN or satisfied in the same sentence as a PENDING id its heading does not name: **two candidates.** This item under REVIEWED-138 — the positive control, found. PENDING-147 under REVIEWED-134 — **a false positive**: *"PENDING-147 leg (i) records preservation discharged 2026-08-19"* describes a preservation run, not the item. Negative control held: REVIEWED-138's own heading id never appears as a candidate. ⚠ **The first instance is invisible to this method by construction** — REVIEWED-132 never names PENDING-139, so there is no text to find. The census rules out further instances only where the ruling names the item.
+
+Recorded at PENDING-146 as well, where the class lives.
+
+**Awaiting:** DISCHARGED — by REVIEWED-138. Closure recorded 2026-09-11 at the steward's direction.
+
 ## PENDING-181 — The paste is the last unremedied transit path, and PENDING-150 routed its remedy here three weeks ago
 
 **Date:** 2026-09-10
@@ -4533,6 +4569,19 @@ Established three ways, none resting on another:
 
 **Awaiting:** unchanged — steward authorization and a jurist design gate on the parent. This addendum narrows the open design point to a recommendation with its alternatives; it decides nothing the parent left to them.
 
+### PENDING-182 — ADDENDUM 2: the design survives the collapse of its stated rationale, and PENDING-184 is a precondition, not a neighbour
+**Date:** 2026-09-11 · filed before any ruling, on the steward's two comments on ADDENDUM 1.
+
+**1 · The field does not depend on the premise ADDENDUM 1 withdrew.** The parent justified the allocation question by concurrent panes; the record shows no two of 746 ticks in one second. **A reader who later falsifies that premise — or confirms it — learns nothing about whether `tick_id` is needed.** The field is needed because a detached child timestamps its own completion, never its tick, and judgment 2's lag partition could not be attributed without a join (REVIEWED-137 §2). That holds with one pane or ten. **The rationale is corrected; the mechanism stands.**
+
+⚠ **The precedent is PENDING-150 §6b, cited for its diagnosis and not its outcome.** §6b's words: *"The mechanics are sound; the justification is false."* There the fix went the other way. The seed rule as filed — `claude/governance/fool/seed/FOOL-SEED-RULE.md`, and `PROVENANCE_COMMIT` in `derive_fool.py` — carries `4d2ae87`, which is §6b's disposition (ii): **the mechanism was changed to honour the rationale.** That was right there, because the rationale — seeding from the constitution as it stood before the fool was conceived — *was* the purpose of choosing a commit. **The test the precedent teaches: when mechanics and justification come apart, ask which one carries the purpose.** In §6b the justification did. Here the mechanism does.
+
+**2 · PENDING-184 is a precondition of this item's usefulness, not a neighbouring item.** Under the id, judgment 2's condition shows as `aside` and `notable` ticks whose child reports late, or never. **A selftest stray that draws `aside` or `notable` is a tick whose child never reports, by construction** — none was spawned. The instrument would inherit a fault that mimics its own finding, and nothing in a tick record lets a reader subtract it afterwards: live and test ticks are indistinguishable there (ADDENDUM 1). ⚠ The two confirmed strays drew `silent`, which a join does not flag. The 27% that draw otherwise would.
+
+**Recommended condition on any ruling that authorizes the field: that it be built only on a clean writer — PENDING-184 closed and its D9/D10 controls green — and not while the contaminant can still write.**
+
+**Awaiting:** steward authorization and a jurist design gate on the parent — **with PENDING-184 closed first.**
+
 ## PENDING-183 — CARRIER: the eleven open items from the verdicts sitting, named so they can be aimed at
 **Date:** 2026-09-10
 **Tag:** [HARDENING]
@@ -4847,3 +4896,28 @@ it fired. **This addendum is that record.**
 **Files affected:** `scripts/tarbuckle-body.py`.
 
 **Awaiting:** nothing further — `[FIX]` lane, steward-directed in session 2026-09-11 (*"Yes, and file it before you build it"*). Filed first; the build follows, and its result is recorded here.
+
+### PENDING-184 — ADDENDUM 1: landed in `37a2c86`, both arms shown to fail by mutation — CLOSED
+**Date:** 2026-09-11
+
+Filed at `2af4335`, fixed at `37a2c86`, in that order. Every run below used a throwaway `HOME`, so no mutant could write a stray:
+
+| run | selftest | fake live log | failed |
+|---|---|---|---|
+| fixed | 34/34 | unchanged | none |
+| pre-fix file, from git | **32/32 green** | **written** | none — the defect's own false pass, on record |
+| M1 — `DRAWS` rebinding deleted | NOT VERIFIED | written | D9, D10 |
+| M2 — `_log_draw`'s write sent to `os.devnull` | NOT VERIFIED | unchanged | **D10 only** |
+| `main()`, tick due, no transcript | prints the mark | **+1 `tick` record** | none |
+
+**Real `HOME`:** 34/34, and the live log's SHA-256, measured outside the selftest, unchanged across the run. **Five-selftest census, re-run:** none of the five writes anything.
+
+**What this establishes.** The D block no longer writes live state. Each arm is shown capable of failing, and **M2 is the case an absence-only control would have certified**: the write vanishes, the live log is untouched, and only D10 notices. **The `main()` run is the control on the fix itself** — the live path still records its tick, so a polluted log was not traded for an empty one.
+
+**What it does not establish.** The past strays stay unrecoverable — a floor of two, left in the log. D9 can fail falsely on a genuine concurrent tick. The fleet's other selftests are unassessed.
+
+**The tag, bound as a claim.** `[FIX]` is the executor's characterization: it addresses the **instance**. The class census was bounded to the five Tarbuckle selftests, and within it the body was the only instance. A fleet-wide census of selftests that write live state would be `[HARDENING]`. It is **not filed**; if wanted, its method is the per-script throwaway-`HOME` census above, and its stated bound is that a script failing under a fake `HOME` is *unassessed*, not clean.
+
+**PENDING-182 ADDENDUM 2's recommended condition — PENDING-184 closed, D9 and D10 green — is met as of `37a2c86`.**
+
+**Awaiting:** CLOSED — `[FIX]` landed in `37a2c86`, steward-directed in session 2026-09-11.