G13 (Helm lane): the pin pill against ground truth -- G10's predicate had no

fresh cell, and no gate asked whether the pill was true

Two findings, measured on six reachable cells of the pill x qualifier pair
(q63/make_cells.py -> R34-CELLS.json), five of them real renders of the real
renderers on the real snapshot and one constructed and labelled as such:

1. G10's predicate (helm-selftest-q62.py:2782) demands both "not re-checked
   since" and "as of ", but helm-render-plain.py:329 appends the disclaimer only
   when site_fresh != "FRESH" -- so the honest fresh generation is refused and
   the end state of G10's own prescribed remedy (re-collect -> re-render ->
   deploy) is red exactly like today's unfixed page. A red that cannot tell a
   fixed page from a broken one is not a remediation signal. The pair is the
   evidence: base gate with the fresh cell as its subject 187/1 with G10's row
   FAIL (control-q62-freshcell.py, c05dd15f), new gate on the same body same
   repoint 190/1 with G10's row PASS (twinD, ef9980cc). The fix splits the
   predicate into its two halves and reads freshness from the age the page's
   own qualifier states, against the collector's own 3600 s threshold.

2. No gate asked whether the pill was TRUE, and G11 does not close it: the
   fresh cells carry an unqualified green "match" beside a fresh-looking clock
   from BOTH the production renderer and the G11-patched one, while a fresh
   fetch of the snapshot's own URL returns HTTP 200 / 217,163 B /
   c34ae2c245117ca7 against a held pin of f27326e8b13d1dec -- so
   pin_matches_live_now is false. G13 is the guard: offline, the pill's colour
   must agree with the held pin against the live-body hash the page publishes;
   live, a fresh fetch must confirm an unqualified green pill (red today, and
   that red is the finding); plus a read-only row. Row 2 subsumes G12 on one
   cell and adds three G12 cannot see -- stated in the record, not left for a
   reader to find.

Gate: base 4b18329c (171,979 B / 3,256 lines / 138 registrations) -> shipped
helm-selftest-q63.py 180,772 B / 3,432 lines / 141 registrations /
94f9b3625e5791a724afe479315f5a49d3244c5e183cad8001ec6185e1c3f240
(+8,793 B / +176 lines / +3 registrations); four twins and one control each one
to three asserted lines different. Subject 4e45021d80395d3f / 29,954 B
throughout. Matrix, prediction stated before the run and met exactly: shipped
189/2 twice, twinA 188/3, twinB 190/1, twinD 190/1, control 187/1, and twinC
rc 0 / 191/0 / ALL GATES PASS -- the honest remediated state, the only
configuration this gate has ever been able to pass, which gives the remediation
an acceptance criterion it did not have before.

Live delta 0 across all seven legs: page 9,865 B / 427b7b54c69b18d2, snapshot
14,039 B / aa842b0fd88037f8, held pin 120 B / 5b6d70716153c3f4; deployed bin
untouched and grep -c HELM_COLLECTOR = 0 so Q40 stands; sandboxes 0/0; disk
41 G free of 154. Nothing deployed, no another-party file edited, decisions/
and inbox/ untouched, no secret printed. Adoption is now four files in one
order: renderer 306d4bee -> collector 4e45021d + gate 94f9b362 -> re-collect ->
re-render -> deploy, with the re-collect load-bearing because G13's live row
stays red until a collect recomputes manifest_match against the live body.

Four defects of mine logged first in the record's section 8: the first probe
fetched the portfolio instead of the pin's subject; the first normaliser missed
this host's TMPDIR-based mkdtemp; the first G13 block referenced four names it
had not imported; and twinC's third repoint silently did not apply until an
assertion caught it. No queue item opened. Merlin <merlin@agents.invalid>.

WORK-QUEUE.md

@@ -9238,3 +9238,121 @@ write). Scratch tree `/root/crucible/q62/`; nothing deployed, no `cp` into `/roo
   the Helm executable set was empty at the start of the run by both this lane's and the pipe's
   measurement, so the unit taken was the one defect the lane had measured and labelled but
   never fixed.
+
+---
+
+**G13 — the pin pill against ground truth (Helm lane; Run 34, 2026-10-15,
+Merlin).** Deliverable `recoveries/helm-pill-ground-truth.md` (**14,419 B /
+sha256 `d6600fd195e4b11c937b6512366242c92269243293c84a28e56c0d51d54b38bb`**, this host, at the write). Scratch tree
+`/root/crucible/q63/`; nothing deployed, no `cp` into `/root/.helm/bin/`, no
+write to the live document root, `decisions/` and `inbox/` untouched, no
+another-party file edited, no secret printed. **No queue item opened** — the
+Helm executable set was empty at the start of the run by both this lane's and
+the pipe's measurement, so the unit taken is the residual G12's own record
+named and left open: *"no fuzzer, the enumerated render set only."*
+
+- **Six reachable cells of the pill x qualifier pair, measured before anything
+  was built** (`q63/make_cells.py` -> `R34-CELLS.json`, 5,261 B / sha256
+  `11d4666a2a587a61…`), five of them real renders of the real renderers on the
+  real snapshot: `plain x stale` 11,806 B / `3f48643d8e7bac77…` (green, both
+  strings); `plain x fresh` 11,624 B / `1eec213e9e363d36…` (green, no
+  disclaimer); `g11 x stale` 11,819 B / `951a8a7645fe7ef4…` (amber, both
+  strings); `g11 x fresh` 11,624 B / `1eec213e9e363d36…` (green, no disclaimer);
+  the deployed page 9,865 B / `427b7b54c69b18d2…` (green, neither string); and
+  `drift x fresh` 11,624 B / `c199045d43ee4006…` (red, no disclaimer), the one
+  **CONSTRUCTED** cell — no snapshot on this host carries a frozen
+  `manifest_match` of false, so the honest remediated state is a scratch renderer
+  copy with one boolean flipped, labelled as constructed everywhere it is cited.
+- **Ground truth, measured in the same run:** held pin 120 B /
+  `5b6d70716153c3f4…`, content `f27326e8b13d1dec…` pinned
+  2026-10-08T16:58:07.100Z; a fresh fetch of the snapshot's own
+  `stackchain_site.url` (`https://www.stackchainmagazine.net/`) returns **HTTP
+  200 / 217,163 B / `c34ae2c245117ca7…`**, so **`pin_matches_live_now` is
+  false** — byte-identical to every reading taken on 2026-10-14.
+- **Finding 1 — G10's predicate has no cell for the honest fresh state.**
+  `helm-selftest-q62.py:2782` demands both `"not re-checked since"` and
+  `"as of "`, and `helm-render-plain.py:329` appends the disclaimer **only when
+  `site_fresh != "FRESH"`**, so the fresh cells are refused and the end state of
+  G10's own prescribed remedy (re-collect -> re-render -> deploy) is **red
+  exactly like today's unfixed page**. A red that cannot tell a fixed page from
+  a broken one is not a remediation signal. **The pair is the evidence:** the
+  base gate with the fresh cell as its subject is **187/1, G10's row FAIL**
+  (`control-q62-freshcell.py`), and the new gate on the same body with the same
+  repoint is **190/1 with G10's row PASS** (`twinD`). The fix splits the
+  predicate into its two halves — the cell states its clock, and says
+  "not re-checked since" exactly when the reading behind it is stale — with
+  freshness parsed from the age the page's own qualifier states, against the
+  same 3600 s threshold the collector uses.
+- **Finding 2 — no gate checked the pill against the pin, and G11 does not close
+  it.** G10 owns the missing clock and G12 owns the incoherent *pair*; neither
+  asks whether the pill is **true**. Cells 2 and 4 carry an unqualified green
+  `match` beside a fresh-looking clock, emitted by **both** the production
+  renderer and the G11-patched one, while the pin does not match the live body.
+  The mechanism is structural: the pill reads the **frozen** `manifest_match`
+  field (`helm-render-plain.py:323`, `:326`) while the qualifier is computed at
+  render time (`:327`-`:330`), and `site_fresh` comes from the render clock,
+  which `--as-of` supplies. **G13 is the fix, three rows:** offline, the pill's
+  colour must agree with the held pin against the live-body hash the page itself
+  publishes (green on all six cells today, so a real check that is currently
+  satisfied); live, a fresh fetch of the snapshot's own URL must confirm an
+  unqualified green pill (**red today, and that red is the finding**); and
+  read-only. **Overlap stated, not hidden:** row 2 subsumes G12 on cell 1 and
+  adds cells 2, 4 and 5, which G12 cannot see; G12 stays load-bearing as the
+  offline half.
+- **The gate.** Base `../q62/helm-selftest-q62.py` 171,979 B / 3,256 lines /
+  138 registrations / `4b18329c7437bec4…` -> shipped
+  `helm-selftest-q63.py` **180,772 B / 3,432 lines / 141 registrations / sha256
+  `94f9b3625e5791a724afe479315f5a49d3244c5e183cad8001ec6185e1c3f240`**
+  (**+8,793 B / +176 lines / +3 registrations**). Twins `f5452b17…` (A),
+  `041a5766…` (B), `a7a5bb1b…` (C), `ef9980cc…` (D), each one to three asserted
+  lines different; control `c05dd15f…` is the base with the same one-line
+  repoint. Subject collector throughout
+  `HELM_COLLECTOR=/root/crucible/q48/helm-collect.py`, 29,954 B /
+  `4e45021d80395d3f03fc58d7e99610e3c1c0d2fcf9c85f1db50b16cfd2043c6c`, asserted
+  by digest before the run.
+- **The matrix, prediction stated before the run and met exactly**
+  (`run_matrix_r34.py` -> `out/R34-GATE.json`, 21,662 B,
+  `generated_at 2026-10-15T03:50:24+00:00`): shipped **189/2** (G10's deployed
+  row and G13's live row) twice; twinA 188/3 (adds G12 row 1); twinB 190/1
+  (G13's rows green on the honest remediated body); **twinC rc 0, 191/0, ALL
+  GATES PASS**; twinD 190/1; control 187/1. **Two results carry the run: twinC
+  is the only configuration the gate has ever been able to pass, and it is the
+  honest remediated state — a fresh clock with a pill that is not claiming a
+  match — so the remediation now has an acceptance criterion that did not exist
+  before; and twinD vs control isolates finding 1's fix and nothing else.**
+- **Live state delta 0** across all seven legs (bytes + sha256 + mtime_ns,
+  `final_r34.py` -> `out/R34-FINAL.json`, `2026-10-15T03:50:56+00:00`): page
+  9,865 B / `427b7b54c69b18d2…` / `1791709034017837233`; snapshot 14,039 B /
+  `aa842b0fd88037f8…` / `1791479579475780303`; held pin 120 B /
+  `5b6d70716153c3f4…` / `1791478687099553071`. Deployed `/root/.helm/bin/`
+  untouched (five digests re-read plus the `__pycache__/` an adoption must not
+  copy in) and `grep -c HELM_COLLECTOR /root/.helm/bin/helm-selftest.py` =
+  **0**, so Q40 stands and none of Q31-G13 is in production. Sandboxes 0/0
+  across `/root/crucible`. Disk 41 G free of 154 (74 %). ctime on the snapshot
+  and the pin reads `1791957230760646867` / `1791957230761646950` — guard
+  restores of earlier runs, disclosed.
+- **Four defects of mine, logged first in the record's §8:** the first probe
+  fetched `https://alexanderwhitestone.com/` — the portfolio, a different
+  artifact from the pin's subject — and read 13,016 B /
+  `b6bae048a2f6bf19…`, which would have made `pin_matches_live_now: false` true
+  for the wrong reason (corrected before anything was recorded; the comparison
+  is in `refetch_r34.py`); the first stdout normaliser carried only the
+  `/tmp/tmp` pattern while this host's TMPDIR is
+  `/root/.hermes/profiles/merlin/cache/scratch`, leaving exactly one line
+  differing between two runs of the same gate (raw stdouts kept in
+  `out/R34-SHIPPED-{1,2}.txt`); the first G13 block referenced `_g13_Path`,
+  `_g13_sh`, `_g13_sys` and `_g13_tf` without importing them and died at line
+  3274 with a `NameError`; and twinC was first derived from twinB, so its third
+  repoint silently did not apply — the assertion
+  `twinC.count("drift-fresh.html") == 3` now catches that class.
+- **Adoption, now four files in one order:** adopt renderer `306d4bee…`
+  (22,731 B) -> adopt collector `4e45021d…` (29,954 B) + gate
+  **`94f9b362…`** (180,772 B, superseding `4b18329c…` and `125754b270bf…`) ->
+  re-collect -> re-render -> deploy. The **re-collect** step is now load-bearing
+  rather than cosmetic: G13's live row stays red until a collect recomputes
+  `manifest_match` against the live body, because the pin does not match it
+  today, so deploying without re-collecting publishes a page this gate refuses.
+- **Standing, unchanged and still human-gated:** that `cp`; the stale
+  `manifest_match` pin (measured false on two instruments); the Q13 publication
+  call; Q9's checksum handover; Q19's onboarding. **No item closed, none
+  opened.**

checkins/2026-10-15-merlin.md

@@ -198,3 +198,99 @@ is why the rule could be shown to fire on something that could actually be
 deployed.
 
 — Merlin (Run 33, closing)
+
+---
+
+### Run 34 — the pin pill against ground truth
+
+**Screen.** Same screen as Run 33, re-measured rather than carried: the queue's
+own hygiene ledger at `:8017` still records **9 opening markers, 8 belonging to
+closed items, 1 genuinely open** (Q52, Mercury's), and every `**Status:` line in
+the file matching an open/unclaimed form names a closed item except **Q9**,
+which its own record has blocked on the audio-handover decision. The Helm
+lane's executable set is empty by both this lane's and the pipe's measurement,
+and everything standing is human-gated. **So the queue offered no finishable
+item and I continued my lane**, discharging the one limit G12's own record
+named and left open: *"no fuzzer, the enumerated render set only."* The run's
+instruments are stamped 03:39 (the cell space), 03:50:24 (the matrix) and
+03:50:56 (the final census) on 2026-10-15.
+
+**Did. Two findings, one predicate fix, one new guard — and the first
+configuration this gate has ever been able to pass.**
+
+- **The cell space, measured before anything was built.** Six reachable cells of
+  the pill x qualifier pair, five of them real renders of the real renderers on
+  the real snapshot (`q63/make_cells.py` -> `R34-CELLS.json`): green+both
+  strings, green+no disclaimer (twice, from the production renderer **and** the
+  G11-patched one), amber+both strings, green+neither string (the deployed
+  page), and red+no disclaimer (the one constructed cell). Ground truth in the
+  same run: a fresh fetch of the snapshot's own URL returns **HTTP 200 /
+  217,163 B / `c34ae2c245117ca7…`** against a held pin of
+  `f27326e8b13d1dec…`, so **`pin_matches_live_now` is false**.
+- **Finding 1 — G10's predicate had no cell for the honest fresh state.**
+  `helm-render-plain.py:329` appends `" · not re-checked since"` only when
+  `site_fresh != "FRESH"`, so a predicate demanding both strings refuses the
+  fresh generation and the end state of G10's own remedy is **red exactly like
+  today's unfixed page**. Evidence, one body two predicates: the base gate with
+  the fresh cell as its subject is **187/1 with G10's row FAIL**
+  (`control-q62-freshcell.py`, 171,975 B / `c05dd15f…`); the new gate on the
+  same body with the same repoint is **190/1 with G10's row PASS** (twinD,
+  180,768 B / `ef9980cc…`). The fix splits the predicate into its two halves
+  and reads freshness from the age the page's own qualifier states, against the
+  collector's own 3600 s threshold.
+- **Finding 2 — no gate asked whether the pill was true, and G11 does not close
+  it.** G10 owns the missing clock, G12 owns the incoherent pair; the fresh
+  cells carry an unqualified green `match` beside a fresh-looking clock while
+  the pin does not match the live body, and G12 pairs the pill only with the
+  disclaimer. **G13 is the guard:** offline, the pill's colour must agree with
+  the held pin against the live-body hash the page publishes; live, a fresh
+  fetch must confirm an unqualified green pill — **red today, and that red is
+  the finding**; plus a read-only row. Row 2 subsumes G12 on one cell and adds
+  three G12 cannot see, which is stated in the record rather than left for a
+  reader to find.
+- **The gate and the matrix.** Base `4b18329c7437bec4…` (171,979 B / 3,256
+  lines / 138 registrations) -> shipped `helm-selftest-q63.py` **180,772 B /
+  3,432 lines / 141 registrations / `94f9b3625e5791a7…`**, +8,793 B / +176
+  lines / +3 registrations, four twins and one control each one to three
+  asserted lines different. Subject `4e45021d80395d3f…` / 29,954 B throughout.
+  Seven configurations, prediction stated before the run and met exactly:
+  shipped **189/2** twice, twinA 188/3, twinB 190/1, twinD 190/1, control
+  187/1 — and **twinC rc 0, 191/0, ALL GATES PASS**, the honest remediated
+  state, the only configuration the gate has ever been able to pass.
+- **Live state delta 0** on bytes + sha256 + mtime_ns across all seven legs:
+  page 9,865 B / `427b7b54c69b18d2…`; snapshot 14,039 B /
+  `aa842b0fd88037f8…`; held pin 120 B / `5b6d70716153c3f4…`. Deployed
+  `/root/.helm/bin/` untouched, `grep -c HELM_COLLECTOR` = **0** so Q40 stands.
+  Sandboxes 0/0. Disk 41 G free of 154. The two shipped runs are byte-identical
+  after normalising the four `mkdtemp` shapes, residual 0.
+- **Four defects of mine, logged first:** the first probe fetched the portfolio
+  instead of the pin's subject and would have made the drift finding true for
+  the wrong reason; the first normaliser missed this host's TMPDIR-based
+  `mkdtemp` (Run 33's own blind spot, reproduced by its author); the first G13
+  block referenced four names it had not imported and died at line 3274; and
+  twinC's third repoint silently did not apply until an assertion caught it.
+
+**Blocked.** The remediation, not the work. G13 detects and cannot repair: the
+remedy is **re-collect -> re-render -> deploy**, an operator's call under §4,
+and the gate is red on today's page for a real and named reason — a green pill a
+fresh fetch refutes. Standing and still human-gated, unchanged and not
+re-argued: the §4 adoption `cp` (now collector `4e45021d80395d3f…` + gate
+**`94f9b3625e5791a7…`**, this run's, superseding `4b18329c…`), the stale
+`manifest_match` pin, the Q13 publication call, Q9's checksum handover, Q19's
+onboarding. Run 33 still awaits the pipe's verification; reply 58 discharged
+Run 32 and carried no ask of this lane.
+
+**Next.** The adoption sequence is four files in one order — renderer
+`306d4bee…`, collector `4e45021d…`, gate `94f9b362…`, then re-collect, re-render,
+deploy — and the **re-collect** is now load-bearing rather than cosmetic,
+because G13's live row stays red until a collect recomputes `manifest_match`
+against the live body. With that done the Helm lane's executable set returns to
+empty and its next unit is again a decision rather than a task.
+
+**Ask.** Alexander: the adoption order, and the acceptance criterion that did
+not exist before this run — the sequence is done when the gate reaches twinC's
+shape, **rc 0 / 191 passed / 0 failed / ALL GATES PASS**, which today's page
+cannot reach and a correctly re-collected page can. Nothing is asked of the
+pipe; no counterparty ask arrived this run.
+
+— Merlin (Run 34)

recoveries/helm-pill-ground-truth.md

@@ -0,0 +1,262 @@
+# The pin pill against ground truth — Run 34
+
+**Deliverable:** this file. **Scratch tree:** `/root/crucible/q63/`. Nothing
+deployed: no `cp` into `/root/.helm/bin/`, no `--deploy`, no write to the live
+document root. `decisions/` and `inbox/` untouched, no another-party file
+edited, no secret printed. **No queue item opened** — the Helm executable set
+was empty at the start of this run by both this lane's and the pipe's
+measurement, so the unit taken is the residual G12's own record named and left
+open: *"no fuzzer, the enumerated render set only."*
+
+**Grade: measured, not claimed.** Every figure below came from a command in
+this run; the instruments are named with their artifacts. Where a figure is a
+dated reading it says so.
+
+---
+
+## 1. The premise, measured before anything was built
+
+`/root/crucible/q63/make_cells.py` → `R34-CELLS.json`. Six reachable cells of
+the pin-pill × qualifier pair, five of them **real renders of the real
+renderers on the real snapshot** and one a constructed fixture, labelled as
+one (§6):
+
+| cell | bytes | sha256 | pill | `as of` | `not re-checked` | G10 | G12 | G13r1 |
+|---|---|---|---|---|---|---|---|---|
+| `plain x stale` | 11,806 | `3f48643d8e7bac77…` | `#22c55e` match | 1 | 1 | PASS | **FAIL** | PASS |
+| `plain x fresh` | 11,624 | `1eec213e9e363d36…` | `#22c55e` match | 1 | 0 | **FAIL** | PASS | PASS |
+| `g11 x stale` | 11,819 | `951a8a7645fe7ef4…` | `#f59e0b` match (unverified) | 1 | 1 | PASS | PASS | PASS |
+| `g11 x fresh` | 11,624 | `1eec213e9e363d36…` | `#22c55e` match | 1 | 0 | **FAIL** | PASS | PASS |
+| `deployed page` | 9,865 | `427b7b54c69b18d2…` | `#22c55e` match | 0 | 0 | **FAIL** | PASS | PASS |
+| `drift x fresh` *(constructed)* | 11,624 | `c199045d43ee4006…` | `#ef4444` drift | 1 | 0 | **FAIL** | PASS | PASS |
+
+Ground truth, measured in the same run (`final_r34.py` → `out/R34-FINAL.json`,
+`generated_at 2026-10-15T03:50:56+00:00`):
+
+- held pin `/root/.helm/data/held/stackchain-home.html.sha256`, 120 B /
+  sha256 `5b6d70716153c3f4bd82045d5d71edf88f793a240ce1275c3dfca5759d0ee297`,
+  content `f27326e8b13d1dec9eb158fe40e5091f79dcaf73b47312c4e7d3958ba1d95d18
+  stackchain-home.html  pinned 2026-10-08T16:58:07.100Z`
+- the snapshot's `stackchain_site`: `manifest_match: true`,
+  `live_body_sha256: f27326e8b13d1dec…`, `observed_at 2026-10-08T17:12:59.458Z`,
+  `url https://www.stackchainmagazine.net/`
+- **a fresh fetch of that URL: HTTP 200 / 217,163 B / sha256
+  `c34ae2c245117ca73ba5720b2c22b6fab689840070ff68cda7bffe3e723986d4`**, so
+  **`pin_matches_live_now` is `false`** — measured, and byte-identical to every
+  reading the pipe and this lane took on 2026-10-14.
+
+**Defect of mine, logged first:** the first version of `probe_r34.py`
+hardcoded `https://alexanderwhitestone.com/` — the portfolio, a *different
+artifact* from the pin's subject — and read 13,016 B /
+`b6bae048a2f6bf19…`, which would have made `pin_matches_live_now: false` true
+for the wrong reason. The URL is now read from the snapshot's own
+`stackchain_site.url`; the comparison is in `refetch_r34.py`. Nothing
+published before that correction.
+
+---
+
+## 2. Finding 1 — G10's predicate has no cell for the honest fresh state
+
+`helm-selftest-q62.py:2782` demands **both** `"not re-checked since"` **and**
+`"as of "`. The honest FRESH generation omits the disclaimer **by
+construction**: `helm-render-plain.py:329` appends `" · not re-checked since"`
+only when `site_fresh != "FRESH"`. So cells 2 and 4 — the real renderers at a
+fresh clock — are refused, and the end state of G10's own prescribed remedy
+(re-collect → re-render → deploy) is **red exactly like today's unfixed
+page**. A red that cannot tell a fixed page from a broken one is not a
+remediation signal.
+
+Measured, and the pair is the evidence (`run_matrix_r34.py`):
+
+| config | G10's deployed row |
+|---|---|
+| `control-q62-freshcell.py` — the **base** gate, subject repointed at the fresh cell | **FAIL** (187/1) |
+| `helm-selftest-q63-twinD.py` — the **new** gate, same body, same repoint | **PASS** (190/1, its one failure is G13's live row) |
+
+Same body, same repoint, one difference: the predicate.
+
+**The fix.** The predicate is two halves, the second conditional: the cell
+states the clock behind its claim, and says `"not re-checked since"` exactly
+when the reading behind it is stale. Freshness is read from the age the page's
+own qualifier states, parsed against the same 3600 s threshold the collector
+uses (`hc.PER_METRIC_STALENESS_S["stackchain_site.http_status"]`), so the gate
+cannot be argued into accepting a page by changing a clock it does not read.
+
+---
+
+## 3. Finding 2 — no gate checked the pill against the pin, and G11 does not close it
+
+G10 owns the missing clock. G12 owns the incoherent *pair*. **Neither asks
+whether the pill is true.** Cells 2 and 4 carry an unqualified green `match`
+beside a fresh-looking clock, emitted by **both** the production renderer and
+the G11-patched one, while the held pin does not match the live body a fresh
+fetch returns. G12's predicate pairs the pill only with the disclaimer, so a
+green pill with a fresh-looking clock passes it.
+
+The mechanism, read in the source rather than inferred: the pill's colour and
+label come from the **frozen** `manifest_match` field
+(`helm-render-plain.py:323`, `:326`) while the qualifier is computed at render
+time (`:327`–`:330`), and `site_fresh` is derived from the render clock —
+which `--as-of` supplies. So a clock can launder a six-day-old reading into a
+live-looking claim, and nothing read the claim against the pin.
+
+**The fix — G13, three rows:**
+
+1. **offline** — the pill's colour must agree with the held pin compared
+   against the live-body hash the page itself publishes. Catches a collector
+   that freezes a claim contradicting the two hashes it printed. **Green on all
+   six cells today** (the published hash *is* the pin), so this is a real check
+   that is currently satisfied, not a vacuous one.
+2. **live** — a fresh fetch of the snapshot's own URL must confirm an
+   unqualified green pill. **Red today, and that red is the finding:** the
+   deployed page claims a match a fresh fetch refutes. If the fetch cannot be
+   made the row is red with the reason named — a green claim that cannot be
+   checked is not a claim a gate may pass. That is Q35's trade, taken
+   knowingly and stated rather than buried.
+3. **read-only** — G13 reads the deployed page and writes nothing.
+
+**Overlap stated, not hidden:** row 2 *subsumes* G12 on cell 1 — both refuse
+it — and adds cells 2, 4 and 5, which G12 cannot see. G12 stays load-bearing
+as the offline half, because row 2 needs the network.
+
+---
+
+## 4. The gate
+
+Base `../q62/helm-selftest-q62.py` **171,979 B / 3,256 lines / 138 `check("`
+registrations / sha256 `4b18329c7437bec4c5c05be86ddd8376a7396eda9c97739410b1ab5952e4d270`**.
+
+| file | bytes | lines | regs | sha256 |
+|---|---|---|---|---|
+| `helm-selftest-q63.py` (shipped) | 180,772 | 3,432 | 141 | `94f9b3625e5791a724afe479315f5a49d3244c5e183cad8001ec6185e1c3f240` |
+| `helm-selftest-q63-twinA.py` | 180,768 | 3,432 | 141 | `f5452b17ef136c65d1eca759d781624dcd7e171d73c82cd389b1d533d0eaad71` |
+| `helm-selftest-q63-twinB.py` | 180,766 | 3,432 | 141 | `041a5766251a6eabb467eb066ade4cb48b26c8ab873a661efb0259e91b6931e3` |
+| `helm-selftest-q63-twinC.py` | 180,760 | 3,432 | 141 | `a7a5bb1b5f3c82d1f08862055dce4da8eb8c0d01868e10c81d6e2c63e4f8d53b` |
+| `helm-selftest-q63-twinD.py` | 180,768 | 3,432 | 141 | `ef9980ccd7fe1ee529068333bdd9a708affa1cdc75a139e8269c252a54ac5ed8` |
+| `control-q62-freshcell.py` | 171,975 | 3,256 | 138 | `c05dd15fac35734e5082734f834440859ccda1399b20137edfddef38d9569b74` |
+
+**+8,793 B / +176 lines / +3 registrations** on the base; every twin is one,
+two or three asserted lines different from the shipped gate, each printed by
+`build_g13.py` (twinA L3107, twinB L3274, twinC L2764+L3107+L3274, twinD L2764,
+control L2764 on the base). Subject collector throughout:
+`HELM_COLLECTOR=/root/crucible/q48/helm-collect.py`,
+**29,954 B / `4e45021d80395d3f03fc58…`** (asserted by digest before the run).
+
+---
+
+## 5. The matrix — prediction stated before the run, met exactly
+
+`run_matrix_r34.py` → `out/R34-GATE.json` (21,662 B,
+`generated_at` in the artifact), seven configurations, live census around each:
+
+| config | rc | passed / failed | the failures |
+|---|---|---|---|
+| `shipped` | 1 | **189 / 2** | G10 deployed row; **G13 live row** |
+| `shipped-again` | 1 | **189 / 2** | the same two — byte-identical after normalisation |
+| `twinA` (G12 subject = `plain-stale`) | 1 | 188 / 3 | the same two **+ G12 row 1** |
+| `twinB` (G13 subject = `g11-stale`) | 1 | 190 / 1 | G10 deployed row only — G13's rows green on the honest remediated body |
+| `twinC` (**all three subjects = `drift-fresh`**) | **0** | **191 / 0** | **ALL GATES PASS** |
+| `twinD` (G10 subject = `plain-fresh`) | 1 | 190 / 1 | G13 live row only |
+| `control` (base gate, old predicate, same body as twinD) | 1 | 187 / 1 | G10 deployed row |
+
+Two results carry the run:
+
+- **twinC reaches ALL GATES PASS, and it is the only configuration that
+  does.** It is the honest remediated state — a fresh clock with a pill that
+  is not claiming a match. The gate has never been able to go green; now it
+  can, and only there. That is the remediation's acceptance criterion, which
+  did not exist before this run.
+- **twinD vs control isolates finding 1's fix and nothing else:** same body,
+  same repoint, old predicate red, new predicate green.
+
+**Determinism:** the two shipped runs are byte-identical after normalising the
+four `mkdtemp` shapes, residual 0. **Defect of mine, logged first:** the first
+normaliser carried only the `/tmp/tmp` pattern, but this host's `TMPDIR` is
+`/root/.hermes/profiles/merlin/cache/scratch`, so `mkdtemp(prefix="tmp")`
+lands at `scratch/tmpXXXXXXXX` and exactly one line (the deploy leg's render
+destination) survived. The same blind spot Run 33 logged against its own
+normaliser, reproduced by its author on a fresh artifact; the raw stdouts are
+kept in `out/R34-SHIPPED-{1,2}.txt` so the residual can be read rather than
+asserted away.
+
+**Live state, delta 0** across all seven legs (`out/R34-FINAL.json`,
+`2026-10-15T03:50:56+00:00`): deployed page 9,865 B /
+`427b7b54c69b18d2…` / mtime_ns `1791709034017837233`; snapshot 14,039 B /
+`aa842b0fd88037f8…` / `1791479579475780303`; held pin 120 B /
+`5b6d70716153c3f4…` / `1791478687099553071`. Deployed `/root/.helm/bin/`
+untouched — five files re-read (`ee0b36a82f23…` / `5c67473d8928…` /
+`7ed1bca8f9da…` / `12570786ba70…` / `738699986778…`) plus the `__pycache__/`
+an adoption must not copy in — and `grep -c HELM_COLLECTOR
+/root/.helm/bin/helm-selftest.py` = **0**, so Q40 stands and none of
+Q31–G13 is in production. Sandboxes **0/0** across `/root/crucible`. Disk 41 G
+free of 154 (74 %). ctime on the snapshot and the pin has moved to
+`1791957230760646867` / `1791957230761646950` (guard restores of earlier runs;
+disclosed, and it is why the run-wide instrument reads ctime as a restore
+trace rather than as proof of a write).
+
+---
+
+## 6. The one constructed cell, and the limits
+
+`drift-fresh.html` is **not** a render of anything on disk. No snapshot the
+collector has written carries a frozen `manifest_match` of `false` — every one
+says `true` — so the honest remediated state cannot be rendered from the live
+data. It is produced by one boolean flip in a **scratch copy** of the renderer
+(`q63/helm-render-drift.py`, `SNAPSHOT` repointed at
+`q63/snapshot-drift.json`), which is Q45's redirect class applied to a
+fixture. It is labelled CONSTRUCTED everywhere it is cited, and twinC's result
+rests on it: **the claim is that the gate can pass that body, not that the body
+exists in production.**
+
+Limits, stated as the item's first limits:
+
+- G13 detects and cannot repair. The remedy is still **re-collect → re-render →
+  deploy**, an operator's call under §4 — but it now has an acceptance
+  criterion (twinC's shape) and a gate that can see whether it landed.
+- Row 2 needs the network. Offline it is red with the reason named, which is
+  the conservative direction and will read as noise on a disconnected host.
+- Row 1 is green on all six cells today, so it has never fired. It is a real
+  check (it compares three published objects), not a demonstrated catch.
+- No fuzzer: the six cells are enumerated, not generated. The pair's space is
+  small and reachable-by-construction, which is why enumeration is honest
+  here, but it is still enumeration.
+- The pill's *substance* is still the collector's frozen field. G13 makes a
+  false green visible; it does not make the field fresh.
+
+---
+
+## 7. Adoption — now four files in one order
+
+The §4 `cp` is unchanged in kind and longer by one object. Order:
+
+**adopt renderer `306d4bee…` (22,731 B) → adopt collector `4e45021d…`
+(29,954 B) + gate `94f9b362…` (180,772 B) → re-collect → re-render → deploy.**
+
+The gate supersedes `4b18329c…` (this run's base, Run 33's) and
+`125754b270bf…` (Q45S10+G10's). The **re-collect** step is now load-bearing
+rather than cosmetic: G13's live row stays red until a collect recomputes
+`manifest_match` against the live body, because the pin does not match it
+today. Deploying without re-collecting publishes a page this gate refuses.
+
+**Standing, unchanged and still human-gated:** that `cp`; the stale
+`manifest_match` pin (measured false, twice, on two instruments); the Q13
+publication call; Q9's checksum handover; Q19's onboarding. **No queue item
+opened.**
+
+---
+
+## 8. Defects of mine, logged first
+
+1. `probe_r34.py` fetched the wrong URL (§1) — caught by this run's own
+   re-read, corrected before anything was recorded, comparison kept in
+   `refetch_r34.py`.
+2. The first normaliser missed this host's TMPDIR-based `mkdtemp` (§5).
+3. The first G13 block referenced `_g13_Path`, `_g13_sh`, `_g13_sys` and
+   `_g13_tf` without importing them; the gate died at line 3274 with a
+   `NameError` on the first run. Caught by running it, not by reading it.
+4. `build_g13.py`'s twinC was first derived from twinB, so its third repoint
+   silently did not apply and the twin differed on two lines rather than the
+   three it claimed. The assertion `twinC.count("drift-fresh.html") == 3` now
+   catches that class; the claim and the artifact agree.
+
+— Merlin (Run 34), Helm lane