diff --git a/persistent-memory.d/2026-09-21-ambiguity-pass-did-the-work.md b/persistent-memory.d/2026-09-21-ambiguity-pass-did-the-work.md
new file mode 100644
index 0000000..4ff8788
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-ambiguity-pass-did-the-work.md
@@ -0,0 +1,11 @@
+# Every code-changing finding came from the AMBIGUITY pass
+
+_2026-09-21 · booth_
+
+**Every one of the panel's code-changing findings came from the
+AMBIGUITY pass, none from a paraphrase divergence** — and two arms independently
+proposed cutting the paraphrase to a drift-check for narrative-heavy contracts,
+because this contract's own frontmatter carries a plain-language narrative and the
+paraphrase was partly reading my framing back to me. That is a finding about the
+`/heid-contract-review` **skill**, not about this repo, and it was reported back
+to heid. Recorded here only so a future session does not rediscover it.
diff --git a/persistent-memory.d/2026-09-21-boolean-escape-hatch-as-lifetime.md b/persistent-memory.d/2026-09-21-boolean-escape-hatch-as-lifetime.md
new file mode 100644
index 0000000..c65482c
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-boolean-escape-hatch-as-lifetime.md
@@ -0,0 +1,11 @@
+# A boolean escape hatch as the lifetime mechanism
+
+_2026-09-21 · booth_
+
+**A boolean escape hatch as the lifetime mechanism.**
+`.forever` was added because a 24h TTL genuinely did not fit some booths —
+and then 56% of live booths ended up on it, which means it is not "ephemeral
+with an exception", it is two lifetimes wearing one lifetime's clothes, with
+the operator doing the sorting by hand. Replaced at U4 by lifetime derived
+from state (an open mark pins; viewing is activity; `keep` survives as an
+explicit reasoned pin rather than the only way to say "not yet").
diff --git a/persistent-memory.d/2026-09-21-deterministic-order-invariant.md b/persistent-memory.d/2026-09-21-deterministic-order-invariant.md
new file mode 100644
index 0000000..6d680bb
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-deterministic-order-invariant.md
@@ -0,0 +1,16 @@
+# Deterministic order is a cross-cutting v1 invariant
+
+_2026-09-21 · booth_
+
+**Deterministic order is a cross-cutting v1 invariant** —
+operator directive, mid-implementation. Every ordered collection the Booth
+renders must have a *stated* rule producing the same sequence on every render
+of the same state; the rule can be anything defensible (byte order, time, an
+explicit number, an arbitrary-but-recorded sequence), but no rule at all is
+forbidden. It binds harder here than elsewhere because the Booth's job is
+**comparison** — the operator judges tile 47 against tile 47 and refers to
+artifacts positionally, so an order that moves between renders misfiles a flag
+or a note rather than crashing. Recorded as `ROADMAP.md` § "Cross-cutting
+invariant" (with the per-collection table) and `CLAUDE.md` invariant 6, and
+tested. Still undecided and must be settled before those units ship: **U7's
+section ordering and compare pairing**, and **U6's bench listing**.
diff --git a/persistent-memory.d/2026-09-21-extracted-from-eshpfi.md b/persistent-memory.d/2026-09-21-extracted-from-eshpfi.md
new file mode 100644
index 0000000..a5a4b94
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-extracted-from-eshpfi.md
@@ -0,0 +1,7 @@
+# Extracted from `eshpfi` into its own repo
+
+_2026-09-21 · booth_
+
+**Extracted from `eshpfi` into its own repo.** The accreted
+service came over whole, tests included, so `tests/test_booth.py` (1581 lines)
+is the regression net the v1 rewrite is checked against.
diff --git a/persistent-memory.d/2026-09-21-five-mechanisms-one-job.md b/persistent-memory.d/2026-09-21-five-mechanisms-one-job.md
new file mode 100644
index 0000000..808103d
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-five-mechanisms-one-job.md
@@ -0,0 +1,13 @@
+# Five mechanisms to get one question beside one artifact
+
+_2026-09-21 · booth_
+
+**Five separate mechanisms to get one question next to one
+artifact** — `.forever`, the link board, `inline.py`'s placeholder DSL,
+`wrap_verbatim_html`'s six regexes, and the floating amber asks chip plus
+`/b/<n>/asks`. Every one is a *correct local fix* to the same global
+mismatch, which is exactly why they accumulated without anyone making a bad
+call. **The foot-gun is the sixth one:** the next "just add a small thing for
+this case" reads as reasonable and is the pattern. The git log carries the
+signature — every feature ships, then takes 2–5 patches for cases the single
+shape did not anticipate. Check the ROADMAP gate before adding a mechanism.
diff --git a/persistent-memory.d/2026-09-21-forever-diagnosis-is-a-prediction.md b/persistent-memory.d/2026-09-21-forever-diagnosis-is-a-prediction.md
new file mode 100644
index 0000000..2361989
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-forever-diagnosis-is-a-prediction.md
@@ -0,0 +1,10 @@
+# The `.forever` diagnosis is a falsifiable prediction
+
+_2026-09-21 · booth_
+
+**The `.forever` diagnosis is a stated, falsifiable
+prediction.** U4 (derived lifetime) predicts the kept-rate falls to the
+genuinely-durable booths. Re-measured today: **14 of 25 booths kept (56%)**,
+against the 54% the IA doc recorded. **Re-count a fortnight after U4 lands.**
+If it does not move, the diagnosis was wrong and the boolean was doing
+something else. Tracked in the IA doc's Booth section and by this entry.
diff --git a/persistent-memory.d/2026-09-21-ia-and-v1-gate-landed.md b/persistent-memory.d/2026-09-21-ia-and-v1-gate-landed.md
new file mode 100644
index 0000000..1876eac
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-ia-and-v1-gate-landed.md
@@ -0,0 +1,11 @@
+# The information architecture and the v1 gate landed
+
+_2026-09-21 · booth_
+
+**The information architecture and the v1 gate landed**
+(`726822b`): `docs/design/information-architecture.md` names the single
+defect — *one lifetime (24h from last touch) and one shape (a folder),
+serving five jobs with different lifetimes and different shapes* — and
+`ROADMAP.md` gates v1 on seven units, each closing a **measured** defect
+rather than a wish. Both were written after a measurement pass over the live
+service, and the measurements are the load-bearing part.
diff --git a/persistent-memory.d/2026-09-21-jinja-hot-reload-outage.md b/persistent-memory.d/2026-09-21-jinja-hot-reload-outage.md
new file mode 100644
index 0000000..1249903
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-jinja-hot-reload-outage.md
@@ -0,0 +1,24 @@
+# Letting Jinja hot-reload templates in the deployment root
+
+_2026-09-21 · booth_
+
+**Letting Jinja hot-reload templates while the repo is the
+deployment root** — the cause of a live outage the same day U2 landed, and the
+sharpest foot-gun in the repo. `booth.service` sets `WorkingDirectory` to this
+repo, so the running service imports these files with no build step and no
+staging copy. Python is read once at process start; Jinja's `FileSystemLoader`
+re-reads a template **on every render**. Editing `booth.html` therefore
+deployed it instantly against Python from 22:03 that knew nothing about
+`item_marks`, and **19 of 25 live booths returned 500** with
+`UndefinedError: 'item_marks' is undefined`. Neither the old code nor the new
+code was broken — the service was running both at once.
+**The lesson that generalises:** a skew between a process and the disk under it
+is invisible to the test suite by construction, so no amount of green tests
+would have caught it; the operator found it. Fixed at the source rather than
+with a reminder — the `Environment` is hand-built with `auto_reload=False`, so
+there is now ONE staleness rule (nothing takes effect until you restart) and
+the running process is always a coherent snapshot of one commit. Asserted by
+`test_templates_do_not_hot_reload_from_disk`. Watch the second-order risk the
+fix introduces: a hand-built `Environment` does not inherit `autoescape` from
+the `Jinja2Templates` constructor, and booth names, item names and mark text
+are all agent-authored strings landing in HTML.
diff --git a/persistent-memory.d/2026-09-21-link-board-absorbing-announce.md b/persistent-memory.d/2026-09-21-link-board-absorbing-announce.md
new file mode 100644
index 0000000..bcab933
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-link-board-absorbing-announce.md
@@ -0,0 +1,12 @@
+# Letting the link board absorb the announce job
+
+_2026-09-21 · booth_
+
+**Letting the link board absorb the announce job.** `booth
+link` is an `O_APPEND` write with no identity and no stated rule, so
+re-announcing a bench appends a row instead of updating one, and a booth URL
+rots the moment its booth is swept — **145 of 211 rows (69%) pointed at
+nothing**, and 22 were the same target re-posted (talk 5×, peedlar 4×). The
+rot is **structural, not drift**. The lesson that cost the most: enforcing
+the link rule without first giving the announce job a home (`.booth.json`
+provenance on the index, U5) just makes it homeless.
diff --git a/persistent-memory.d/2026-09-21-marks-storage-decision.md b/persistent-memory.d/2026-09-21-marks-storage-decision.md
new file mode 100644
index 0000000..345939a
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-marks-storage-decision.md
@@ -0,0 +1,21 @@
+# Marks are one `.marks.json` per booth
+
+_2026-09-21 · booth_
+
+**Marks are stored as one `.marks.json` per booth**, atomic
+temp-file + `os.replace`, `fcntl` lock on the read-modify-write — operator
+decision, this session. Two alternatives were weighed and lost: a sidecar
+per item (`<rel>.marks.json`) and extending the existing `<stem>.ask.json`
+shape. Rationale, and the reason it is not `links.md`-shaped: **(a)** U4
+makes *"does this booth owe an answer?"* a hot question — the sweep asks it
+per booth per tick and the index asks it per card per page load, so per-item
+sidecars turn it into a full walk of all 25 booths, one of which holds 270
+files; **(b)** `links.md` is an `O_APPEND` content-hash log because **17
+agent handles write it concurrently**, whereas marks have exactly one writer
+(the operator, in one browser) and many readers — a different problem that
+must not inherit the append-log design; **(c)** `.blurred` / `.pins` /
+`.forever` already establish the per-booth dotfile as the house shape for
+operator state, and `booth_items()`'s dotfile skip means it costs nothing in
+counts, galleries or zips. Accepted cost: a corrupt `.marks.json` loses that
+booth's marks rather than one item's. Implementation deferred to U2 —
+tracked at `ROADMAP.md` U2 and by this entry.
diff --git a/persistent-memory.d/2026-09-21-marks-write-wiped-judgment.md b/persistent-memory.d/2026-09-21-marks-write-wiped-judgment.md
new file mode 100644
index 0000000..c22b631
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-marks-write-wiped-judgment.md
@@ -0,0 +1,15 @@
+# A write over a damaged `.marks.json` wiped the booth
+
+_2026-09-21 · booth_
+
+**A write over a damaged `.marks.json` was wiping every mark in
+the booth.** Shipped in `v0.2.0`, found by the panel (Kimi, converged with
+Hulda), fixed in `v0.2.1`. `marks_for` is deliberately lenient — unparseable
+reads as `[]` so a review page still loads — and the write path inherited that
+leniency through the same reader, so one flag click appended to an empty list and
+atomically replaced the file. The fix is an **asymmetry**, which is the reusable
+part: reads stay lenient, writes go strict (`MarksCorrupt`), damaged bytes stay
+on disk, routes answer 409 not 500. A page that renders without an annotation is
+recoverable; a file that overwrote the operator's judgment is not. Kimi also
+named the class correctly — "an author steeped in the design conversation would
+likely read past" it — and that was accurate.
diff --git a/persistent-memory.d/2026-09-21-partial-answer-counts-as-open.md b/persistent-memory.d/2026-09-21-partial-answer-counts-as-open.md
new file mode 100644
index 0000000..5f0c344
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-partial-answer-counts-as-open.md
@@ -0,0 +1,10 @@
+# A partially-answered pick counts as OPEN
+
+_2026-09-21 · booth_
+
+**A partially-answered pick now counts as OPEN** — declared, not
+smuggled. The old index badge tested `answer is None`, so a half-answered
+four-question ask read as closed on the index while the panel beside it
+rendered `◐ partial`: the two disagreed about the same booth. Open is the
+reading that makes U4 correct — a lifetime rule that unpinned a booth on the
+first radio click would sweep a review in flight.
diff --git a/persistent-memory.d/2026-09-21-regex-injecting-chrome.md b/persistent-memory.d/2026-09-21-regex-injecting-chrome.md
new file mode 100644
index 0000000..63c8f5a
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-regex-injecting-chrome.md
@@ -0,0 +1,14 @@
+# Regex-injecting chrome into arbitrary author HTML
+
+_2026-09-21 · booth_
+
+**Regex-injecting chrome into arbitrary author HTML**
+(`wrap_verbatim_html` + `_HEAD_CLOSE_RE`, `_HTML_OPEN_RE`, `_DOCTYPE_RE`,
+`_BODY_CLOSE_RE`, `_HTML_CLOSE_RE`, `_ICON_RE`, and the doctype/charset
+ordering constraints they thread). It works today and is **still live** —
+but it is the single most fragile thing in the service and it is load-bearing
+for the operator's most important workflow. Slated for deletion at U3 in
+favour of a declared seam (`/_booth/embed.js`, mounted through a real DOM
+API), which costs an author one line and removes the whole class. Do not
+extend the regex set in the meantime; if a verbatim page breaks, that is an
+argument for U3, not for a seventh pattern.
diff --git a/persistent-memory.d/2026-09-21-sindra-finalists-is-the-motivation.md b/persistent-memory.d/2026-09-21-sindra-finalists-is-the-motivation.md
new file mode 100644
index 0000000..cb12ffa
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-sindra-finalists-is-the-motivation.md
@@ -0,0 +1,8 @@
+# `sindra-finalists` is U2's flag motivation, caught live
+
+_2026-09-21 · booth_
+
+**`sindra-finalists` is U2's `flag` motivation caught in the
+act** — 86 items, every one captioned, and the booth's entire name is "the
+ones the operator picked." That loop currently runs through chat, which is
+the defect `flag` closes. Evidence, not argument.
diff --git a/persistent-memory.d/2026-09-21-tagging-with-a-gate-in-flight.md b/persistent-memory.d/2026-09-21-tagging-with-a-gate-in-flight.md
new file mode 100644
index 0000000..a5c19a5
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-tagging-with-a-gate-in-flight.md
@@ -0,0 +1,12 @@
+# Tagging a release while a review gate was in flight
+
+_2026-09-21 · booth_
+
+**Tagging a release while a review gate was still in flight.**
+`v0.2.0` was cut and announced to 15 consuming handles; the
+`/heid-contract-review` panel — dispatched BEFORE implementation, as the
+discipline says — replied afterwards with three defects in the code that had just
+shipped, one of them silent data loss. Nothing about the tier decision was wrong;
+the *timing* was. **If a gate is outstanding on the work being released, the tag
+waits for it.** The cost was a same-hour `v0.2.1` and a correction note to peers
+who had already verified against the broken version.
diff --git a/persistent-memory.d/2026-09-21-tolerant-writer-over-tolerant-reader.md b/persistent-memory.d/2026-09-21-tolerant-writer-over-tolerant-reader.md
new file mode 100644
index 0000000..7b6a1cd
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-tolerant-writer-over-tolerant-reader.md
@@ -0,0 +1,9 @@
+# Letting the write path share the read path's leniency
+
+_2026-09-21 · booth_
+
+**Letting the write path share the read path's leniency.** See the
+`MarksCorrupt` decision above. The general shape, worth carrying beyond marks:
+a tolerant reader and a tolerant writer over the same state are not the same
+decision, and pointing both at one function silently makes them one. Tolerate on
+read so the surface still renders; refuse on write so nothing is destroyed.
diff --git a/persistent-memory.d/2026-09-21-two-gates-are-complementary.md b/persistent-memory.d/2026-09-21-two-gates-are-complementary.md
new file mode 100644
index 0000000..d6d04e0
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-two-gates-are-complementary.md
@@ -0,0 +1,13 @@
+# Seam review and cold panel had zero overlap, twice
+
+_2026-09-21 · booth_
+
+**The two review gates are complementary, measured on one unit.**
+The caller-side **seam review** (nine findings, against the real sibling module
+surfaces) and the cold **`/heid-contract-review` panel** (four arms,
+artifact-only) had **zero overlap in both directions** on U2. The seam review
+found a scope miss the panel structurally could not see: the contract omitted
+`inline.py`, whose `place()` indexes by subscript, which a frozen dataclass
+refuses. The panel found three code defects and a missing test the seam review
+had no lens for. Matches heid's kvasir zero-overlap result on the
+conformance-versus-hunt axis. **Run both; neither substitutes.**
diff --git a/persistent-memory.d/2026-09-21-u2-marks-landed.md b/persistent-memory.d/2026-09-21-u2-marks-landed.md
new file mode 100644
index 0000000..927d1e4
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-u2-marks-landed.md
@@ -0,0 +1,16 @@
+# U2 (marks) landed — one primitive for three mechanisms
+
+_2026-09-21 · booth_
+
+**U2 (marks) landed.** One primitive replacing three
+mechanisms. `pick` / `note` / `flag` in one `.marks.json` per booth, one read
+path (`marks_for`), one openness predicate (`open_marks`), rendered beside the
+artifact on the tile, at full size in the zoom, and in the panel. `flag` and
+`note` had no write path at all before this — the selection loop
+(`golden-candidates`, `sindra-finalists`, the `pancake-*` ladders) was running
+through chat. 242 tests. Details worth carrying: `asks.py` kept `normalize_ask`
+and gained `build_answer` (the 2026-09-09 partial-answer semantics preserved by
+moving, not rewriting) and LOST its five sidecar-storage functions;
+`GET /b/<n>/marks.json` was added because remote sessions polled
+`<stem>.answer.json` over HTTP and the sidecar's removal would have taken that
+capability with it; `/b/<n>/asks` 308s to `/marks`.
diff --git a/persistent-memory.d/2026-09-21-u2-seam-review-earned-it.md b/persistent-memory.d/2026-09-21-u2-seam-review-earned-it.md
new file mode 100644
index 0000000..77f9751
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-u2-seam-review-earned-it.md
@@ -0,0 +1,17 @@
+# The U2 seam review earned its place, and how
+
+_2026-09-21 · booth_
+
+**The U2 seam review earned its place, and the record should
+say how.** Nine findings against the real `booth.asks` / `booth.items` /
+`booth.inline` surfaces, two of which changed scope or behaviour: `inline.py`
+was missing from `touches` entirely (its `place()` indexes asks by
+**subscript**, which a frozen dataclass refuses — nothing else in the service
+does that), and the partial-answer inconsistency above. The cold
+`/heid-contract-review` pass is artifact-only by design and structurally
+cannot see a sibling module, so neither it nor a same-model self-review would
+have found either. Two more surfaced later and are worth the same note: a
+SECOND subscript in `inline.place` the seam review undercounted, and a
+regression in my own legacy importer that a retargeted test caught — a
+malformed sidecar that renders `⚠ broken` today would have silently vanished
+on migration.
diff --git a/persistent-memory.d/2026-09-21-u7-section-premise-half-wrong.md b/persistent-memory.d/2026-09-21-u7-section-premise-half-wrong.md
new file mode 100644
index 0000000..b7c72be
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-u7-section-premise-half-wrong.md
@@ -0,0 +1,19 @@
+# U7's section premise is half wrong
+
+_2026-09-21 · booth_
+
+**U7's section premise is half wrong, and it is the half that
+matters** — found by re-measuring `~/booth-data` rather than trusting the IA
+doc. The IA says sections come from subfolders that already exist on disk;
+true, but **every booth that actually needs navigation is flat**:
+`pancake-v3-full` (270 items, 0 subfolders), `pancake-v4-full` (270, 0),
+`sindra20-engines` (98 items + 99 caption sidecars, 0), `sindra-finalists`
+(86 + 87, 0). Subfolders exist on exactly two booths — `pewpew-ui-brief` (7,
+nested to `_ds/powerpellet-design-system-<uuid>/preview`) and `dfa-concepts`
+(1) — and **both are reports**, the job where grid navigation matters least.
+So sections stay worth shipping and `Item.section` stays right, but they are
+**not** "most of the navigation fix": the rail, the filters and grid keyboard
+are all of it. Worth noting for whoever writes U7: `sindra20-engines` encodes
+its structure in the **filename prefix** (`b2-s1-<subject>-<seed>`), which is
+where a grouping heuristic would actually pay. The IA doc's claim about what
+sections buy needs a line struck — not yet edited.
diff --git a/persistent-memory.d/2026-09-21-v020-tagged-with-a-gate-in-flight.md b/persistent-memory.d/2026-09-21-v020-tagged-with-a-gate-in-flight.md
new file mode 100644
index 0000000..7aa747e
--- /dev/null
+++ b/persistent-memory.d/2026-09-21-v020-tagged-with-a-gate-in-flight.md
@@ -0,0 +1,14 @@
+# v0.2.0 was tagged while a gate was in flight
+
+_2026-09-21 · booth_
+
+**v0.2.0 cut and announced; v0.2.1 fixed what the announcement
+was already wrong about.** Operator approved the minor (a v1 unit closed plus a
+CLI surface change for 17 consuming handles clears the release-note bar). The
+note went to 15 handles — the 17 link-board posters minus `nh3-dev`, a host
+label, and `heid`, an oracle that does not script these verbs. Then the
+cross-frontier contract panel landed and found **three defects in the code I had
+just released**, so `v0.2.1` shipped within the hour. Sequence worth remembering:
+the release was correct by the tier bar and still premature by the discipline —
+the panel had been dispatched BEFORE implementation and its reply arrived AFTER
+the tag. **If a gate is in flight, the tag can wait for it.**
diff --git a/persistent-memory.d/2026-09-22-cli-exit-codes.md b/persistent-memory.d/2026-09-22-cli-exit-codes.md
new file mode 100644
index 0000000..75f83ce
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-cli-exit-codes.md
@@ -0,0 +1,14 @@
+# `booth marks` / `booth answer` got real exit codes
+
+_2026-09-22 · booth_
+
+**`booth marks` / `booth answer` got real exit codes**, because
+a read that CRASHED was indistinguishable from a read that said no. `marks`
+printed a traceback and exited 0 (a caller's `jq` saw success and got
+nothing); `answer --wait` read a damaged file as "not yet" and spun for the
+full hour before blaming the operator. Now `0 ok · 1 unanswered/timed-out ·
+2 no such pick · 3 unreadable`, and `read_error()` was added to `marks.py` so
+the CLI can ask the question the browser must not: the page stays lenient, the
+machine consumer gets the truth. Also `--wait` now prints ONCE — it was
+emitting a whole JSON document per poll, so a captured `--wait` held several
+concatenated values and parsed as none of them.
diff --git a/persistent-memory.d/2026-09-22-doctrine-not-defect.md b/persistent-memory.d/2026-09-22-doctrine-not-defect.md
new file mode 100644
index 0000000..65bd11f
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-doctrine-not-defect.md
@@ -0,0 +1,13 @@
+# An existing test stopped me retiring documented behaviour
+
+_2026-09-22 · booth_
+
+**An existing test stopped me retiring documented behaviour
+while fixing a race.** The mtime-restore race is real, and the clean fix —
+ignoring a booth directory's own mtime whenever the booth holds anything —
+would also have silently retired the rule that RELEASING a kept board resets
+its clock, which the CLI header, the README and a deliberately-written test
+all pin. That is a TTL doctrine change, not a bug fix. Fixed the concrete half
+(a failing `os.utime` used to escape and 500 the route), left the race stated
+in the code. **A fix that changes a documented rule is a proposal, not a
+patch.**
diff --git a/persistent-memory.d/2026-09-22-lenient-reader-blast-radius.md b/persistent-memory.d/2026-09-22-lenient-reader-blast-radius.md
new file mode 100644
index 0000000..1e4cb0c
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-lenient-reader-blast-radius.md
@@ -0,0 +1,15 @@
+# The lenient reader's blast radius was the whole service
+
+_2026-09-22 · booth_
+
+**The lenient reader's blast radius was the whole service, not
+one booth.** `_clean_text` did `(text or "").replace(...)` and `marks_for`
+sorts on `(created, id)`, so a stored `text` that was a dict or a `created`
+that was a number raised out of the READ path — and `list_booths` reads every
+booth's marks on every index load. One hand-edited file 500'd `/` and
+`/healthz` for all 25 booths. Fixed in two layers, matching the house posture:
+a named type check (`_entry_type_error`) plus a `_hydrate_safe` backstop that
+cannot raise, and the panel now RENDERS an unreadable mark as ⚠ broken instead
+of as an empty note. **The general shape: a lenient reader is only lenient if
+the leniency is bounded by where it runs.** `marks_for` was written for one
+booth's page and is called in a loop over every booth.
diff --git a/persistent-memory.d/2026-09-22-scripts-booth-got-tests.md b/persistent-memory.d/2026-09-22-scripts-booth-got-tests.md
new file mode 100644
index 0000000..1619d98
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-scripts-booth-got-tests.md
@@ -0,0 +1,11 @@
+# `scripts/booth` went from zero tests to five
+
+_2026-09-22 · booth_
+
+**`scripts/booth` had zero tests and now has five**
+(`tests/test_cli.py`). The panel's guard-strength tables returned UNVERIFIED
+for every CLI claim because nothing in the suite executed the script — two of
+the round's findings lived in exactly that gap. The new tests run the real
+script under the system `python3`, which makes them a live check on INV-1
+(stdlib-only) as a side effect: a third-party import in `marks.py` now fails
+in the suite the same way it would fail on a fleet host.
diff --git a/persistent-memory.d/2026-09-22-size-cap-opened-a-hang.md b/persistent-memory.d/2026-09-22-size-cap-opened-a-hang.md
new file mode 100644
index 0000000..13e4bd8
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-size-cap-opened-a-hang.md
@@ -0,0 +1,23 @@
+# The size cap opened a service-wide hang
+
+_2026-09-22 · booth_
+
+**The U5 bug-hunt panel found a service-wide hang that the
+SIZE CAP ITSELF opened — two hours after I added the cap.** `stat` reports
+size 0 for a FIFO and 0 for a symlink to `/dev/zero`, so both sail under a
+byte cap and then `read_text` blocks with no EOF or allocates until the kernel
+intervenes. `list_booths` reads every booth on every `GET /`, so ONE such file
+stalls the front page for the whole service with no error and no recovery
+short of a restart. Reproduced (`timeout` returned 124), fixed with an
+`S_ISREG` check BEFORE the size check in both modules, verified live: the
+index answered 200 in 36 ms with two FIFOs planted. **The reusable shape:
+`st_size` answers a different question than "can this be read", and a bound
+that trusts it inherits everything it does not mean — a hardening fix opened
+a worse hole than the one it closed.** Also adopted: the upload path wrote the
+manifest ABOVE its own cleanup guard (4/4), so a failure orphaned a half-booth
+whose uniquely-named leaked temp then kept it alive forever; replace-over-
+damaged destroyed recoverable bytes (4/4, now QUARANTINED rather than refused
+— marks refuse because judgment is not restatable, a booth's description is);
+and `booth answer` spelled out its own openness test, disagreeing with
+`booth marks` about a partially-answered pick, which is a direct violation of
+U2's INV-2. Full triage in `persistent-memory.d/2026-09-22-u5-panels.md`.
diff --git a/persistent-memory.d/2026-09-22-u2-bug-hunt-panel.md b/persistent-memory.d/2026-09-22-u2-bug-hunt-panel.md
new file mode 100644
index 0000000..f202acb
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-u2-bug-hunt-panel.md
@@ -0,0 +1,19 @@
+# The U2 bug-hunt panel was not ceremony
+
+_2026-09-22 · booth_
+
+**The U2 bug-hunt panel landed and it was not ceremony —
+`v0.2.2`.** Nine adopted findings across four arms; eight were real against
+live code and one was already fixed. The headline was **4/4 convergent from
+four different angles**: `_Locked.__exit__` unlinked `.marks.lock` on the no-op
+path, and `flock` binds to an INODE — so a writer blocked on the old inode
+proceeds while the next writer creates a fresh lock file and takes it at once.
+Two processes then run the read-modify-write concurrently and the later
+`os.replace` drops a mark, with both of them obeying the protocol. **The
+cleanup existed to protect the booth's TTL and it was failing at that too**:
+creating and removing a directory entry bumps the DIRECTORY's mtime, which is
+what `_newest_mtime` actually seeds from, so a no-op reset the clock it was
+written to leave alone. Same code region, two defects, one fix — never unlink
+the lock, exempt `.<name>.lock` dotfiles from `_newest_mtime`, and put the
+directory's mtime back after creating one. Full triage in
+`persistent-memory.d/2026-09-22-bug-hunt-panel.md`.
diff --git a/persistent-memory.d/2026-09-22-u5-adoption-split-in-two.md b/persistent-memory.d/2026-09-22-u5-adoption-split-in-two.md
new file mode 100644
index 0000000..3077a6b
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-u5-adoption-split-in-two.md
@@ -0,0 +1,23 @@
+# U5's adoption prediction split in two
+
+_2026-09-22 · booth_
+
+**U5's adoption prediction, SPLIT IN TWO within an hour of
+landing — and the split is the interesting part.** The baseline was recorded as
+0 of 26. Fifty minutes after the deploy, `comfy-dev` created `muse-clothed-repro`
+and it announced itself: `{handle: comfy-dev, why: "", created: ...}`. That peer
+was told nothing. **The HANDLE propagates for free** — it rides on `booth new`
+and `booth add`, so every existing CLI caller starts announcing without learning
+anything, which is the flags-on-existing-verbs decision paying off on day zero.
+**The WHY does not** — it needs someone to know the flag exists, and this first
+one is empty.
+
+So re-measure BOTH on **2026-09-29**, because they answer different questions:
+
+    find ~/booth-data -maxdepth 2 -name .booth.json | wc -l          # free
+    grep -l '"why": "[^"]' ~/booth-data/*/.booth.json 2>/dev/null | wc -l   # learned
+
+A high first count and a near-zero second is the predicted shape of "nobody was
+told", and it is the case the operator's no-announcement decision was designed
+to be able to see. Do not read the n=1 above as a rate — it is a code-path
+observation (every CLI caller writes a handle), not a sample.
diff --git a/persistent-memory.d/2026-09-22-u5-panels-reached-a-released-bug.md b/persistent-memory.d/2026-09-22-u5-panels-reached-a-released-bug.md
new file mode 100644
index 0000000..aa8d30c
--- /dev/null
+++ b/persistent-memory.d/2026-09-22-u5-panels-reached-a-released-bug.md
@@ -0,0 +1,21 @@
+# Two U5 panels, and prose reached a released outage
+
+_2026-09-22 · booth_
+
+**Two cross-frontier panels on U5, and a paraphrase panel reached
+a production outage two modules away.** 3-of-4 flagged the contract's "4 GB"
+case as letter-compliant but purpose-defeating; the conformance round found that
+unbounded read live in U5's code; walking it to the sibling found the SAME hole
+**live in released `v0.2.2`** — `marks._read_raw` catches `(OSError, ValueError,
+UnicodeDecodeError)` and `json.loads` on deep nesting raises **RecursionError**,
+which is none of them, so 400 KB of brackets in one booth returned 500 for `/`
+and `/healthz` across all 26. The v0.2.2 round HAD flagged it and I closed half:
+**a finding with two call sites is not closed when one is.** The reusable
+instruction — **walk a conformance finding to the sibling module even when the
+sibling is out of scope.** Five of ten conformance findings were tests of mine
+that pass on the regression they exist to catch, three of them asserting an
+ARTIFACT of the property rather than the property; that is three nights running
+on the same shape. Two real bugs neither my tests nor I could see: a bare
+`booth add` wiped the `why` on the one sequence the feature exists for, and
+`--title` was write-only. Full triage in
+`persistent-memory.d/2026-09-22-u5-panels.md`.
diff --git a/persistent-memory.md b/persistent-memory.md
index 557e872..0d09d9a 100644
--- a/persistent-memory.md
+++ b/persistent-memory.md
@@ -21,352 +21,73 @@ _As of 2026-09-22:_
 
 - **v1 is gated on seven units** in `ROADMAP.md`, dependency-ordered
   **U1 → U2 → {U3, U4, U5} → U7**, with **U6 independent**.
-- **U1 and U2 are landed and released.** Current version `0.2.2`, deployed to the
-  live service, 275 tests green, tree clean, 25/25 booth pages verified 200 after
-  the deploy. U1 `ce598b3`; U2 `c7f9437` released as `v0.2.0`, then `5e41108` as
-  `v0.2.1` (four contract-panel findings), then `v0.2.2` carrying the
-  **bug-hunt panel's** nine (below).
-- **U5 is IMPLEMENTED and unreleased** as of 2026-09-22. `booth/manifest.py`
-  (stdlib-only, INV-1), `.booth.json` per booth, the provenance line on both
-  index lanes and the booth page header, `--why` / `--title` on `booth new` and
-  `booth add`, and the link board + pickup booths announcing themselves as the
-  service's own. 310 tests, live service restarted, 26/26 booth pages verified
-  200 and all 26 rendering `unannounced`. **Deliberately NOT tagged yet**: the
-  cold `/heid-contract-review` panel is still in flight and the code-review and
-  bug-hunt gates have not run. That ordering is the 2026-09-21 lesson applied —
-  a release whose gate is outstanding is premature even when the tier is right.
-  Contract: `docs/contracts/u5_booth_manifest.contract.md` (carries its own
-  seam-review section).
-- **U5's original framing** (operator, 2026-09-21): **self-announcing booths.**
-  `.booth.json` carrying `{handle, title, why, created}`, written by the CLI from
-  `$ALTHING_HANDLE`; the index card gains provenance and a one-line purpose, and
-  the index becomes the "what landed" feed the link board was being used as. It
-  closes job 5 of the five jobs — the one nobody named, and the reason 145 dead
-  link rows existed. Nothing started: no contract, no blast-radius pass.
-- **Two things about U5 are already settled and should not be re-derived.**
-  (1) `.booth.json` is a DOTFILE, so `booth_items`' existing `startswith(".")` skip
-  already keeps it out of tiles, counts and zips — the same reason `.marks.json`
-  needed no new exclusion rule. (2) The deterministic-order invariant applies to
-  whatever U5 adds to the index; the index is ordered newest-first by mtime today
-  and that rule must stay stated. Also worth knowing before scoping: enforcing the
-  link rule without giving job 5 a home first just makes it homeless — that is the
-  lesson from the 69% rot, and U5 is the home.
-- **No heid dispatch is outstanding.** The `/heid-bug-hunt` on U2's diff landed
-  2026-09-22 and shipped as `v0.2.2`; see the dated entry below.
-- Live service `active` on `:8090`, 25 booths, verified 25 × 3 page types after the
-  last deploy. The booth set churns: `sindra20-engines` and `sindra-finalists` were
-  swept during the session, `cr123a-to-d-sleeve` and `sindra` appeared.
+- **U1, U2 and U5 are landed and released.** Current version `0.3.0` (`95beede`,
+  tagged), deployed, 341 tests green, tree clean, 24/24 booth pages verified 200
+  and a full layout-probe sweep at 0 occlusions across 27 pages. U1 `ce598b3`;
+  U2 `c7f9437` → `v0.2.0`, `5e41108` → `v0.2.1`, `026a1fc` → `v0.2.2`;
+  U5 `c015a91` + `95beede` → `v0.3.0`.
+- **THE NEXT UNIT IS THE OPERATOR'S CALL and has not been made.** The session's
+  recommendation is **U4 (derived lifetime)**, on three grounds, none of which
+  need re-deriving: (1) it is the only remaining unit whose defect was
+  re-measured this session — the `.forever` rate is **17 of 24 (70%)**, up from
+  the 54% in the IA doc and the 56% measured on 2026-09-21, and the absolute
+  count rose too (14 → 17), so it is not only the denominator shrinking;
+  (2) U2 built `open_marks` FOR it and says so in the predicate's docstring;
+  (3) the bug-hunt panel's mtime-restore finding was deliberately left open
+  because closing it properly is a TTL doctrine change — and U4 is the unit that
+  owns TTL doctrine, so it closes a loop this session opened on purpose.
+  U3 and U6 are the alternatives: U6 is independent and was *conceptually*
+  unblocked by U5 giving job 5 a home, U3 is where verbatim-booth provenance was
+  deferred to.
+- **No gate is outstanding.** All three ran on U5 and were folded in: the cold
+  `/heid-contract-review` panel (thread `01M340PNVRS21HPASZT38PXQPN`), the
+  `/heid-code-review` conformance panel (`01M341E9XAPZEFBSPK9HPGAM0S`) and the
+  `/heid-bug-hunt` (`01M343SXX27Z47C3STXXRC7M42`). All three loops are closed
+  with heid.
+- **Two dated predictions are pending and must not be forgotten.** U5's adoption
+  re-measure on **2026-09-29** (two counts, see its entry — already at 3 of 24
+  announced and 2 with a `why`, all from peers told nothing), and the `.forever`
+  re-count **a fortnight after U4 lands**, which is U4's success criterion.
+- **Three methodology proposals from this session sit with the operator**, routed
+  by heid rather than decided unilaterally: reshaping the paraphrase gate toward
+  a drift-check for narrative-heavy contracts, a standing
+  "green-tests-prove-nothing" direction for the code-review gate, and regin's
+  table-vs-signature consistency pass. They are changes to the `/heid*` skills,
+  not to this repo.
+- The booth set churns hard: 26 → 24 during this session as the sweeper ran.
+  Re-count rather than trusting any number written here.
 
 ## Recent decisions
 
-- `[2026-09-22]` **The U5 bug-hunt panel found a service-wide hang that the
-  SIZE CAP ITSELF opened — two hours after I added the cap.** `stat` reports
-  size 0 for a FIFO and 0 for a symlink to `/dev/zero`, so both sail under a
-  byte cap and then `read_text` blocks with no EOF or allocates until the kernel
-  intervenes. `list_booths` reads every booth on every `GET /`, so ONE such file
-  stalls the front page for the whole service with no error and no recovery
-  short of a restart. Reproduced (`timeout` returned 124), fixed with an
-  `S_ISREG` check BEFORE the size check in both modules, verified live: the
-  index answered 200 in 36 ms with two FIFOs planted. **The reusable shape:
-  `st_size` answers a different question than "can this be read", and a bound
-  that trusts it inherits everything it does not mean — a hardening fix opened
-  a worse hole than the one it closed.** Also adopted: the upload path wrote the
-  manifest ABOVE its own cleanup guard (4/4), so a failure orphaned a half-booth
-  whose uniquely-named leaked temp then kept it alive forever; replace-over-
-  damaged destroyed recoverable bytes (4/4, now QUARANTINED rather than refused
-  — marks refuse because judgment is not restatable, a booth's description is);
-  and `booth answer` spelled out its own openness test, disagreeing with
-  `booth marks` about a partially-answered pick, which is a direct violation of
-  U2's INV-2. Full triage in `persistent-memory.d/2026-09-22-u5-panels.md`.
-- `[2026-09-22]` **An existing test stopped me retiring documented behaviour
-  while fixing a race.** The mtime-restore race is real, and the clean fix —
-  ignoring a booth directory's own mtime whenever the booth holds anything —
-  would also have silently retired the rule that RELEASING a kept board resets
-  its clock, which the CLI header, the README and a deliberately-written test
-  all pin. That is a TTL doctrine change, not a bug fix. Fixed the concrete half
-  (a failing `os.utime` used to escape and 500 the route), left the race stated
-  in the code. **A fix that changes a documented rule is a proposal, not a
-  patch.**
-- `[2026-09-22]` **Two cross-frontier panels on U5, and a paraphrase panel reached
-  a production outage two modules away.** 3-of-4 flagged the contract's "4 GB"
-  case as letter-compliant but purpose-defeating; the conformance round found that
-  unbounded read live in U5's code; walking it to the sibling found the SAME hole
-  **live in released `v0.2.2`** — `marks._read_raw` catches `(OSError, ValueError,
-  UnicodeDecodeError)` and `json.loads` on deep nesting raises **RecursionError**,
-  which is none of them, so 400 KB of brackets in one booth returned 500 for `/`
-  and `/healthz` across all 26. The v0.2.2 round HAD flagged it and I closed half:
-  **a finding with two call sites is not closed when one is.** The reusable
-  instruction — **walk a conformance finding to the sibling module even when the
-  sibling is out of scope.** Five of ten conformance findings were tests of mine
-  that pass on the regression they exist to catch, three of them asserting an
-  ARTIFACT of the property rather than the property; that is three nights running
-  on the same shape. Two real bugs neither my tests nor I could see: a bare
-  `booth add` wiped the `why` on the one sequence the feature exists for, and
-  `--title` was write-only. Full triage in
-  `persistent-memory.d/2026-09-22-u5-panels.md`.
-- `[2026-09-22]` **U5's adoption prediction, SPLIT IN TWO within an hour of
-  landing — and the split is the interesting part.** The baseline was recorded as
-  0 of 26. Fifty minutes after the deploy, `comfy-dev` created `muse-clothed-repro`
-  and it announced itself: `{handle: comfy-dev, why: "", created: ...}`. That peer
-  was told nothing. **The HANDLE propagates for free** — it rides on `booth new`
-  and `booth add`, so every existing CLI caller starts announcing without learning
-  anything, which is the flags-on-existing-verbs decision paying off on day zero.
-  **The WHY does not** — it needs someone to know the flag exists, and this first
-  one is empty.
-
-  So re-measure BOTH on **2026-09-29**, because they answer different questions:
-
-      find ~/booth-data -maxdepth 2 -name .booth.json | wc -l          # free
-      grep -l '"why": "[^"]' ~/booth-data/*/.booth.json 2>/dev/null | wc -l   # learned
-
-  A high first count and a near-zero second is the predicted shape of "nobody was
-  told", and it is the case the operator's no-announcement decision was designed
-  to be able to see. Do not read the n=1 above as a rate — it is a code-path
-  observation (every CLI caller writes a handle), not a sample.
-- `[2026-09-22]` **The U2 bug-hunt panel landed and it was not ceremony —
-  `v0.2.2`.** Nine adopted findings across four arms; eight were real against
-  live code and one was already fixed. The headline was **4/4 convergent from
-  four different angles**: `_Locked.__exit__` unlinked `.marks.lock` on the no-op
-  path, and `flock` binds to an INODE — so a writer blocked on the old inode
-  proceeds while the next writer creates a fresh lock file and takes it at once.
-  Two processes then run the read-modify-write concurrently and the later
-  `os.replace` drops a mark, with both of them obeying the protocol. **The
-  cleanup existed to protect the booth's TTL and it was failing at that too**:
-  creating and removing a directory entry bumps the DIRECTORY's mtime, which is
-  what `_newest_mtime` actually seeds from, so a no-op reset the clock it was
-  written to leave alone. Same code region, two defects, one fix — never unlink
-  the lock, exempt `.<name>.lock` dotfiles from `_newest_mtime`, and put the
-  directory's mtime back after creating one. Full triage in
-  `persistent-memory.d/2026-09-22-bug-hunt-panel.md`.
-- `[2026-09-22]` **The lenient reader's blast radius was the whole service, not
-  one booth.** `_clean_text` did `(text or "").replace(...)` and `marks_for`
-  sorts on `(created, id)`, so a stored `text` that was a dict or a `created`
-  that was a number raised out of the READ path — and `list_booths` reads every
-  booth's marks on every index load. One hand-edited file 500'd `/` and
-  `/healthz` for all 25 booths. Fixed in two layers, matching the house posture:
-  a named type check (`_entry_type_error`) plus a `_hydrate_safe` backstop that
-  cannot raise, and the panel now RENDERS an unreadable mark as ⚠ broken instead
-  of as an empty note. **The general shape: a lenient reader is only lenient if
-  the leniency is bounded by where it runs.** `marks_for` was written for one
-  booth's page and is called in a loop over every booth.
-- `[2026-09-22]` **`booth marks` / `booth answer` got real exit codes**, because
-  a read that CRASHED was indistinguishable from a read that said no. `marks`
-  printed a traceback and exited 0 (a caller's `jq` saw success and got
-  nothing); `answer --wait` read a damaged file as "not yet" and spun for the
-  full hour before blaming the operator. Now `0 ok · 1 unanswered/timed-out ·
-  2 no such pick · 3 unreadable`, and `read_error()` was added to `marks.py` so
-  the CLI can ask the question the browser must not: the page stays lenient, the
-  machine consumer gets the truth. Also `--wait` now prints ONCE — it was
-  emitting a whole JSON document per poll, so a captured `--wait` held several
-  concatenated values and parsed as none of them.
-- `[2026-09-22]` **`scripts/booth` had zero tests and now has five**
-  (`tests/test_cli.py`). The panel's guard-strength tables returned UNVERIFIED
-  for every CLI claim because nothing in the suite executed the script — two of
-  the round's findings lived in exactly that gap. The new tests run the real
-  script under the system `python3`, which makes them a live check on INV-1
-  (stdlib-only) as a side effect: a third-party import in `marks.py` now fails
-  in the suite the same way it would fail on a fleet host.
-
-- `[2026-09-21]` **v0.2.0 cut and announced; v0.2.1 fixed what the announcement
-  was already wrong about.** Operator approved the minor (a v1 unit closed plus a
-  CLI surface change for 17 consuming handles clears the release-note bar). The
-  note went to 15 handles — the 17 link-board posters minus `nh3-dev`, a host
-  label, and `heid`, an oracle that does not script these verbs. Then the
-  cross-frontier contract panel landed and found **three defects in the code I had
-  just released**, so `v0.2.1` shipped within the hour. Sequence worth remembering:
-  the release was correct by the tier bar and still premature by the discipline —
-  the panel had been dispatched BEFORE implementation and its reply arrived AFTER
-  the tag. **If a gate is in flight, the tag can wait for it.**
-- `[2026-09-21]` **A write over a damaged `.marks.json` was wiping every mark in
-  the booth.** Shipped in `v0.2.0`, found by the panel (Kimi, converged with
-  Hulda), fixed in `v0.2.1`. `marks_for` is deliberately lenient — unparseable
-  reads as `[]` so a review page still loads — and the write path inherited that
-  leniency through the same reader, so one flag click appended to an empty list and
-  atomically replaced the file. The fix is an **asymmetry**, which is the reusable
-  part: reads stay lenient, writes go strict (`MarksCorrupt`), damaged bytes stay
-  on disk, routes answer 409 not 500. A page that renders without an annotation is
-  recoverable; a file that overwrote the operator's judgment is not. Kimi also
-  named the class correctly — "an author steeped in the design conversation would
-  likely read past" it — and that was accurate.
-- `[2026-09-21]` **The two review gates are complementary, measured on one unit.**
-  The caller-side **seam review** (nine findings, against the real sibling module
-  surfaces) and the cold **`/heid-contract-review` panel** (four arms,
-  artifact-only) had **zero overlap in both directions** on U2. The seam review
-  found a scope miss the panel structurally could not see: the contract omitted
-  `inline.py`, whose `place()` indexes by subscript, which a frozen dataclass
-  refuses. The panel found three code defects and a missing test the seam review
-  had no lens for. Matches heid's kvasir zero-overlap result on the
-  conformance-versus-hunt axis. **Run both; neither substitutes.**
-- `[2026-09-21]` **Every one of the panel's code-changing findings came from the
-  AMBIGUITY pass, none from a paraphrase divergence** — and two arms independently
-  proposed cutting the paraphrase to a drift-check for narrative-heavy contracts,
-  because this contract's own frontmatter carries a plain-language narrative and the
-  paraphrase was partly reading my framing back to me. That is a finding about the
-  `/heid-contract-review` **skill**, not about this repo, and it was reported back
-  to heid. Recorded here only so a future session does not rediscover it.
-- `[2026-09-21]` **Deterministic order is a cross-cutting v1 invariant** —
-  operator directive, mid-implementation. Every ordered collection the Booth
-  renders must have a *stated* rule producing the same sequence on every render
-  of the same state; the rule can be anything defensible (byte order, time, an
-  explicit number, an arbitrary-but-recorded sequence), but no rule at all is
-  forbidden. It binds harder here than elsewhere because the Booth's job is
-  **comparison** — the operator judges tile 47 against tile 47 and refers to
-  artifacts positionally, so an order that moves between renders misfiles a flag
-  or a note rather than crashing. Recorded as `ROADMAP.md` § "Cross-cutting
-  invariant" (with the per-collection table) and `CLAUDE.md` invariant 6, and
-  tested. Still undecided and must be settled before those units ship: **U7's
-  section ordering and compare pairing**, and **U6's bench listing**.
-- `[2026-09-21]` **U2 (marks) landed.** One primitive replacing three
-  mechanisms. `pick` / `note` / `flag` in one `.marks.json` per booth, one read
-  path (`marks_for`), one openness predicate (`open_marks`), rendered beside the
-  artifact on the tile, at full size in the zoom, and in the panel. `flag` and
-  `note` had no write path at all before this — the selection loop
-  (`golden-candidates`, `sindra-finalists`, the `pancake-*` ladders) was running
-  through chat. 242 tests. Details worth carrying: `asks.py` kept `normalize_ask`
-  and gained `build_answer` (the 2026-09-09 partial-answer semantics preserved by
-  moving, not rewriting) and LOST its five sidecar-storage functions;
-  `GET /b/<n>/marks.json` was added because remote sessions polled
-  `<stem>.answer.json` over HTTP and the sidecar's removal would have taken that
-  capability with it; `/b/<n>/asks` 308s to `/marks`.
-- `[2026-09-21]` **A partially-answered pick now counts as OPEN** — declared, not
-  smuggled. The old index badge tested `answer is None`, so a half-answered
-  four-question ask read as closed on the index while the panel beside it
-  rendered `◐ partial`: the two disagreed about the same booth. Open is the
-  reading that makes U4 correct — a lifetime rule that unpinned a booth on the
-  first radio click would sweep a review in flight.
-- `[2026-09-21]` **The U2 seam review earned its place, and the record should
-  say how.** Nine findings against the real `booth.asks` / `booth.items` /
-  `booth.inline` surfaces, two of which changed scope or behaviour: `inline.py`
-  was missing from `touches` entirely (its `place()` indexes asks by
-  **subscript**, which a frozen dataclass refuses — nothing else in the service
-  does that), and the partial-answer inconsistency above. The cold
-  `/heid-contract-review` pass is artifact-only by design and structurally
-  cannot see a sibling module, so neither it nor a same-model self-review would
-  have found either. Two more surfaced later and are worth the same note: a
-  SECOND subscript in `inline.place` the seam review undercounted, and a
-  regression in my own legacy importer that a retargeted test caught — a
-  malformed sidecar that renders `⚠ broken` today would have silently vanished
-  on migration.
-- `[2026-09-21]` **Marks are stored as one `.marks.json` per booth**, atomic
-  temp-file + `os.replace`, `fcntl` lock on the read-modify-write — operator
-  decision, this session. Two alternatives were weighed and lost: a sidecar
-  per item (`<rel>.marks.json`) and extending the existing `<stem>.ask.json`
-  shape. Rationale, and the reason it is not `links.md`-shaped: **(a)** U4
-  makes *"does this booth owe an answer?"* a hot question — the sweep asks it
-  per booth per tick and the index asks it per card per page load, so per-item
-  sidecars turn it into a full walk of all 25 booths, one of which holds 270
-  files; **(b)** `links.md` is an `O_APPEND` content-hash log because **17
-  agent handles write it concurrently**, whereas marks have exactly one writer
-  (the operator, in one browser) and many readers — a different problem that
-  must not inherit the append-log design; **(c)** `.blurred` / `.pins` /
-  `.forever` already establish the per-booth dotfile as the house shape for
-  operator state, and `booth_items()`'s dotfile skip means it costs nothing in
-  counts, galleries or zips. Accepted cost: a corrupt `.marks.json` loses that
-  booth's marks rather than one item's. Implementation deferred to U2 —
-  tracked at `ROADMAP.md` U2 and by this entry.
-- `[2026-09-21]` **U7's section premise is half wrong, and it is the half that
-  matters** — found by re-measuring `~/booth-data` rather than trusting the IA
-  doc. The IA says sections come from subfolders that already exist on disk;
-  true, but **every booth that actually needs navigation is flat**:
-  `pancake-v3-full` (270 items, 0 subfolders), `pancake-v4-full` (270, 0),
-  `sindra20-engines` (98 items + 99 caption sidecars, 0), `sindra-finalists`
-  (86 + 87, 0). Subfolders exist on exactly two booths — `pewpew-ui-brief` (7,
-  nested to `_ds/powerpellet-design-system-<uuid>/preview`) and `dfa-concepts`
-  (1) — and **both are reports**, the job where grid navigation matters least.
-  So sections stay worth shipping and `Item.section` stays right, but they are
-  **not** "most of the navigation fix": the rail, the filters and grid keyboard
-  are all of it. Worth noting for whoever writes U7: `sindra20-engines` encodes
-  its structure in the **filename prefix** (`b2-s1-<subject>-<seed>`), which is
-  where a grouping heuristic would actually pay. The IA doc's claim about what
-  sections buy needs a line struck — not yet edited.
-- `[2026-09-21]` **`sindra-finalists` is U2's `flag` motivation caught in the
-  act** — 86 items, every one captioned, and the booth's entire name is "the
-  ones the operator picked." That loop currently runs through chat, which is
-  the defect `flag` closes. Evidence, not argument.
-- `[2026-09-21]` **The information architecture and the v1 gate landed**
-  (`726822b`): `docs/design/information-architecture.md` names the single
-  defect — *one lifetime (24h from last touch) and one shape (a folder),
-  serving five jobs with different lifetimes and different shapes* — and
-  `ROADMAP.md` gates v1 on seven units, each closing a **measured** defect
-  rather than a wish. Both were written after a measurement pass over the live
-  service, and the measurements are the load-bearing part.
-- `[2026-09-21]` **The `.forever` diagnosis is a stated, falsifiable
-  prediction.** U4 (derived lifetime) predicts the kept-rate falls to the
-  genuinely-durable booths. Re-measured today: **14 of 25 booths kept (56%)**,
-  against the 54% the IA doc recorded. **Re-count a fortnight after U4 lands.**
-  If it does not move, the diagnosis was wrong and the boolean was doing
-  something else. Tracked in the IA doc's Booth section and by this entry.
-- `[2026-09-21]` **Extracted from `eshpfi` into its own repo.** The accreted
-  service came over whole, tests included, so `tests/test_booth.py` (1581 lines)
-  is the regression net the v1 rewrite is checked against.
+- `[2026-09-22]` **The size cap opened a service-wide hang** — a FIFO has st_size 0; a bound that trusts it inherits what it does not mean → `persistent-memory.d/2026-09-22-size-cap-opened-a-hang.md`
+- `[2026-09-22]` **An existing test stopped me retiring documented behaviour** — the clean fix for the mtime race would have silently changed TTL doctrine → `persistent-memory.d/2026-09-22-doctrine-not-defect.md`
+- `[2026-09-22]` **Two U5 panels, and prose reached a released outage** — read the detail before assuming a conformance finding stops at its own module → `persistent-memory.d/2026-09-22-u5-panels-reached-a-released-bug.md`
+- `[2026-09-22]` **U5's adoption prediction split in two** — the handle rides for free, the why must be learned — RE-MEASURE 2026-09-29 → `persistent-memory.d/2026-09-22-u5-adoption-split-in-two.md`
+- `[2026-09-22]` **The U2 bug-hunt panel was not ceremony** — the lock-unlink race and the TTL guard that was failing at its own job → `persistent-memory.d/2026-09-22-u2-bug-hunt-panel.md`
+- `[2026-09-22]` **The lenient reader's blast radius was the whole service** — marks_for runs per booth per index load; a raise there is an outage → `persistent-memory.d/2026-09-22-lenient-reader-blast-radius.md`
+- `[2026-09-22]` **`booth marks` / `booth answer` got real exit codes** — read it before changing anything the 17 consuming handles call → `persistent-memory.d/2026-09-22-cli-exit-codes.md`
+- `[2026-09-22]` **`scripts/booth` went from zero tests to five** — they run the real script under system python3, so they also check INV-1 → `persistent-memory.d/2026-09-22-scripts-booth-got-tests.md`
+- `[2026-09-21]` **v0.2.0 was tagged while a gate was in flight** — the sequencing lesson: if a gate is outstanding, the tag waits → `persistent-memory.d/2026-09-21-v020-tagged-with-a-gate-in-flight.md`
+- `[2026-09-21]` **A write over a damaged `.marks.json` wiped the booth** — the reads-lenient / writes-strict asymmetry, and why it exists → `persistent-memory.d/2026-09-21-marks-write-wiped-judgment.md`
+- `[2026-09-21]` **Seam review and cold panel had zero overlap, twice** — evidence for running both; neither substitutes for the other → `persistent-memory.d/2026-09-21-two-gates-are-complementary.md`
+- `[2026-09-21]` **Every code-changing finding came from the AMBIGUITY pass** — a finding about the /heid-contract-review skill, not about this repo → `persistent-memory.d/2026-09-21-ambiguity-pass-did-the-work.md`
+- `[2026-09-21]` **Deterministic order is a cross-cutting v1 invariant** — operator directive; read before adding ANY ordered surface → `persistent-memory.d/2026-09-21-deterministic-order-invariant.md`
+- `[2026-09-21]` **U2 (marks) landed — one primitive for three mechanisms** — what moved where, and the HTTP mirror remote sessions poll → `persistent-memory.d/2026-09-21-u2-marks-landed.md`
+- `[2026-09-21]` **A partially-answered pick counts as OPEN** — declared, not smuggled; it is the reading that makes U4 correct → `persistent-memory.d/2026-09-21-partial-answer-counts-as-open.md`
+- `[2026-09-21]` **The U2 seam review earned its place, and how** — inline.place indexes by subscript — the miss a cold panel cannot see → `persistent-memory.d/2026-09-21-u2-seam-review-earned-it.md`
+- `[2026-09-21]` **Marks are one `.marks.json` per booth** — operator decision with two rejected alternatives; read before restructuring → `persistent-memory.d/2026-09-21-marks-storage-decision.md`
+- `[2026-09-21]` **U7's section premise is half wrong** — every booth that needs navigation is FLAT — read before starting U7 → `persistent-memory.d/2026-09-21-u7-section-premise-half-wrong.md`
+- `[2026-09-21]` **`sindra-finalists` is U2's flag motivation, caught live** — evidence, not argument → `persistent-memory.d/2026-09-21-sindra-finalists-is-the-motivation.md`
+- `[2026-09-21]` **The information architecture and the v1 gate landed** — the single defect the seven units decompose → `persistent-memory.d/2026-09-21-ia-and-v1-gate-landed.md`
+- `[2026-09-21]` **The `.forever` diagnosis is a falsifiable prediction** — U4's success criterion — re-count a fortnight AFTER U4 lands → `persistent-memory.d/2026-09-21-forever-diagnosis-is-a-prediction.md`
+- `[2026-09-21]` **Extracted from `eshpfi` into its own repo** — test_booth.py is the regression net the v1 rewrite is checked against → `persistent-memory.d/2026-09-21-extracted-from-eshpfi.md`
 
 ## Tried and abandoned
 
-- `[2026-09-21]` **Tagging a release while a review gate was still in flight.**
-  `v0.2.0` was cut and announced to 15 consuming handles; the
-  `/heid-contract-review` panel — dispatched BEFORE implementation, as the
-  discipline says — replied afterwards with three defects in the code that had just
-  shipped, one of them silent data loss. Nothing about the tier decision was wrong;
-  the *timing* was. **If a gate is outstanding on the work being released, the tag
-  waits for it.** The cost was a same-hour `v0.2.1` and a correction note to peers
-  who had already verified against the broken version.
-- `[2026-09-21]` **Letting the write path share the read path's leniency.** See the
-  `MarksCorrupt` decision above. The general shape, worth carrying beyond marks:
-  a tolerant reader and a tolerant writer over the same state are not the same
-  decision, and pointing both at one function silently makes them one. Tolerate on
-  read so the surface still renders; refuse on write so nothing is destroyed.
-- `[2026-09-21]` **Letting Jinja hot-reload templates while the repo is the
-  deployment root** — the cause of a live outage the same day U2 landed, and the
-  sharpest foot-gun in the repo. `booth.service` sets `WorkingDirectory` to this
-  repo, so the running service imports these files with no build step and no
-  staging copy. Python is read once at process start; Jinja's `FileSystemLoader`
-  re-reads a template **on every render**. Editing `booth.html` therefore
-  deployed it instantly against Python from 22:03 that knew nothing about
-  `item_marks`, and **19 of 25 live booths returned 500** with
-  `UndefinedError: 'item_marks' is undefined`. Neither the old code nor the new
-  code was broken — the service was running both at once.
-  **The lesson that generalises:** a skew between a process and the disk under it
-  is invisible to the test suite by construction, so no amount of green tests
-  would have caught it; the operator found it. Fixed at the source rather than
-  with a reminder — the `Environment` is hand-built with `auto_reload=False`, so
-  there is now ONE staleness rule (nothing takes effect until you restart) and
-  the running process is always a coherent snapshot of one commit. Asserted by
-  `test_templates_do_not_hot_reload_from_disk`. Watch the second-order risk the
-  fix introduces: a hand-built `Environment` does not inherit `autoescape` from
-  the `Jinja2Templates` constructor, and booth names, item names and mark text
-  are all agent-authored strings landing in HTML.
-
-- `[2026-09-21]` **Five separate mechanisms to get one question next to one
-  artifact** — `.forever`, the link board, `inline.py`'s placeholder DSL,
-  `wrap_verbatim_html`'s six regexes, and the floating amber asks chip plus
-  `/b/<n>/asks`. Every one is a *correct local fix* to the same global
-  mismatch, which is exactly why they accumulated without anyone making a bad
-  call. **The foot-gun is the sixth one:** the next "just add a small thing for
-  this case" reads as reasonable and is the pattern. The git log carries the
-  signature — every feature ships, then takes 2–5 patches for cases the single
-  shape did not anticipate. Check the ROADMAP gate before adding a mechanism.
-- `[2026-09-21]` **Regex-injecting chrome into arbitrary author HTML**
-  (`wrap_verbatim_html` + `_HEAD_CLOSE_RE`, `_HTML_OPEN_RE`, `_DOCTYPE_RE`,
-  `_BODY_CLOSE_RE`, `_HTML_CLOSE_RE`, `_ICON_RE`, and the doctype/charset
-  ordering constraints they thread). It works today and is **still live** —
-  but it is the single most fragile thing in the service and it is load-bearing
-  for the operator's most important workflow. Slated for deletion at U3 in
-  favour of a declared seam (`/_booth/embed.js`, mounted through a real DOM
-  API), which costs an author one line and removes the whole class. Do not
-  extend the regex set in the meantime; if a verbatim page breaks, that is an
-  argument for U3, not for a seventh pattern.
-- `[2026-09-21]` **A boolean escape hatch as the lifetime mechanism.**
-  `.forever` was added because a 24h TTL genuinely did not fit some booths —
-  and then 56% of live booths ended up on it, which means it is not "ephemeral
-  with an exception", it is two lifetimes wearing one lifetime's clothes, with
-  the operator doing the sorting by hand. Replaced at U4 by lifetime derived
-  from state (an open mark pins; viewing is activity; `keep` survives as an
-  explicit reasoned pin rather than the only way to say "not yet").
-- `[2026-09-21]` **Letting the link board absorb the announce job.** `booth
-  link` is an `O_APPEND` write with no identity and no stated rule, so
-  re-announcing a bench appends a row instead of updating one, and a booth URL
-  rots the moment its booth is swept — **145 of 211 rows (69%) pointed at
-  nothing**, and 22 were the same target re-posted (talk 5×, peedlar 4×). The
-  rot is **structural, not drift**. The lesson that cost the most: enforcing
-  the link rule without first giving the announce job a home (`.booth.json`
-  provenance on the index, U5) just makes it homeless.
+- `[2026-09-21]` **Tagging a release while a review gate was in flight** — cost a same-hour v0.2.1 and a correction to 15 handles → `persistent-memory.d/2026-09-21-tagging-with-a-gate-in-flight.md`
+- `[2026-09-21]` **Letting the write path share the read path's leniency** — a tolerant reader and a tolerant writer are not the same decision → `persistent-memory.d/2026-09-21-tolerant-writer-over-tolerant-reader.md`
+- `[2026-09-21]` **Letting Jinja hot-reload templates in the deployment root** — caused a live outage: 19 of 25 booths at 500. Why auto_reload=False → `persistent-memory.d/2026-09-21-jinja-hot-reload-outage.md`
+- `[2026-09-21]` **Five mechanisms to get one question beside one artifact** — the accretion signature this whole v1 rewrite is undoing → `persistent-memory.d/2026-09-21-five-mechanisms-one-job.md`
+- `[2026-09-21]` **Regex-injecting chrome into arbitrary author HTML** — the defect U3 exists to close → `persistent-memory.d/2026-09-21-regex-injecting-chrome.md`
+- `[2026-09-21]` **A boolean escape hatch as the lifetime mechanism** — why `.forever` is a symptom; the defect U4 exists to close → `persistent-memory.d/2026-09-21-boolean-escape-hatch-as-lifetime.md`
+- `[2026-09-21]` **Letting the link board absorb the announce job** — 69% rot; U5 gave the job a home, which is what unblocks U6 → `persistent-memory.d/2026-09-21-link-board-absorbing-announce.md`