docs: a fresh bootstrap writes both pairs equal, not a missing converged_*

The previous commit's test and README described a freshly bootstrapped
machine as lacking converged_* and rendering 'not recorded'. That pins the
wrong contract: #61 states plainly that "on a fresh machine both pairs are
written with equal values", and rule 2 only suppresses converged_* churn on
a later same-version re-run.

So two identical lines mean bootstrapped-and-never-re-converged, and a
manifest missing the pair is partial or hand-edited rather than fresh --
no writer produces it. The graceful degradation stays for that case, still
without backfilling from birth, which would invent a convergence that never
happened.

The reader itself was already correct; this fixes the fixture and the prose
that claimed otherwise.

Found in review of #74.

Refs #64

CHANGELOG.md

@@ -33,7 +33,10 @@ on the way to cutting its first release, and this file starts there.
   nothing else depends on it. The reader is keyed to #61's documented schema
   (`schema`, `bootstrapped_by`/`_at`, `converged_by`/`_at`) and fixtures pin
   that exact spelling, so the integration cannot land silently broken; birth
-  and latest stay separate rather than one being inferred from the other. Named `platform` and not `status` on purpose:
+  and latest stay separate rather than one being inferred from the other. A
+  fresh machine writes both pairs equal, so two identical lines read as
+  "never re-converged"; a manifest missing the pair is partial rather than
+  fresh, and says so instead of backfilling from birth. Named `platform` and not `status` on purpose:
   `users status` and `runner status` cross-check recorded against live state
   and print `DRIFT`, and a command that records nothing cannot drift — which
   also leaves `rig status` free for the machine-wide roll-up. Known

README.md

@@ -774,10 +774,19 @@ bootstrapped` on every machine; nothing else in the command depends on it.
 is birth (`bootstrapped_by`/`bootstrapped_at`, first-write-wins, pinned
 forever) and `CONVERGED` is latest (`converged_by`/`converged_at`, updated only
 when the converging version actually differs). That distinction is what answers
-"is this machine still converged by a rig that predates the fix?" — so
-`CONVERGED` reads `not recorded` rather than being backfilled from birth on a
-box that has only ever been bootstrapped once. A manifest whose `schema=` this
-rig does not know is named as such instead of being half-read in silence.
+"is this machine still converged by a rig that predates the fix?".
+
+A **fresh machine writes both pairs with equal values**, so two identical lines
+mean "bootstrapped and never re-converged since" — not a missing record. A
+later re-converge by a different rig moves `CONVERGED` and leaves `BOOTSTRAP`
+untouched, which is the whole point of keeping them apart.
+
+`CONVERGED  not recorded` therefore does **not** describe a freshly
+bootstrapped box; no writer produces a manifest without the pair. It means the
+file is partial or hand-edited, and the value is deliberately not backfilled
+from `BOOTSTRAP` — inferring a convergence that never happened would be worse
+than saying so. Likewise a manifest whose `schema=` this rig does not know is
+named as such instead of being half-read in silence.
 
 **Known limitation — `CPU` and `MEMORY` inside a container-style guest are
 unverified.** `CPU` and `MEMORY` are read straight from `/proc/cpuinfo` and

test/cli.sh

@@ -1023,12 +1023,20 @@ check "platform: reads #61's converged_by/at" 0 "CONVERGED  0.6.0, 2026-08-02T09
   env RIG_MANIFEST="$PLATWORK/manifest" RIG_ROLE_MARKER="$PLATWORK/absent" "$ROOT/bin/rig" platform
 check "platform: reads #61's bootstrapped_by/at" 0 "BOOTSTRAP  0.4.0, 2026-07-19T14:24:51Z" \
   env RIG_MANIFEST="$PLATWORK/manifest" RIG_ROLE_MARKER="$PLATWORK/absent" "$ROOT/bin/rig" platform
-# Birth and latest stay SEPARATE: #61 rule 2 writes converged_* only when the
-# version actually differs, so its absence on a freshly bootstrapped box is a
-# legitimate state — it must not be silently backfilled from bootstrapped_*.
-printf 'schema=1\nbootstrapped_by=0.4.0\nbootstrapped_at=2026-07-19T14:24:51Z\n' > "$PLATWORK/manifest-birth"
-check "platform: unconverged box says so, never infers from birth" 0 "CONVERGED  not recorded" \
-  env RIG_MANIFEST="$PLATWORK/manifest-birth" RIG_ROLE_MARKER="$PLATWORK/absent" "$ROOT/bin/rig" platform
+# A FRESH bootstrap carries both pairs with EQUAL values — #61 is explicit
+# ("On a fresh machine both pairs are written with equal values"); rule 2 only
+# suppresses converged_* churn on a later same-version re-run. So equal dates
+# are the never-re-converged case and must render as themselves, not be
+# special-cased into looking unset.
+printf 'schema=1\nbootstrapped_by=0.4.0\nbootstrapped_at=2026-07-19T14:24:51Z\nconverged_by=0.4.0\nconverged_at=2026-07-19T14:24:51Z\n' > "$PLATWORK/manifest-fresh"
+check "platform: a fresh bootstrap shows both pairs equal (#61)" 0 "CONVERGED  0.4.0, 2026-07-19T14:24:51Z" \
+  env RIG_MANIFEST="$PLATWORK/manifest-fresh" RIG_ROLE_MARKER="$PLATWORK/absent" "$ROOT/bin/rig" platform
+# A manifest missing converged_* is therefore NOT a fresh box — no writer
+# produces that — so it is partial or hand-edited. Degrade loudly rather than
+# backfilling from birth, which would invent a convergence that never happened.
+printf 'schema=1\nbootstrapped_by=0.4.0\nbootstrapped_at=2026-07-19T14:24:51Z\n' > "$PLATWORK/manifest-partial"
+check "platform: a partial manifest says so, never infers from birth" 0 "CONVERGED  not recorded" \
+  env RIG_MANIFEST="$PLATWORK/manifest-partial" RIG_ROLE_MARKER="$PLATWORK/absent" "$ROOT/bin/rig" platform
 # A newer schema renders what it recognises and says the rest is unreadable,
 # rather than pretending a partial read is the whole truth.
 printf 'schema=2\nbootstrapped_by=9.9.9\nbootstrapped_at=2027-01-01T00:00:00Z\n' > "$PLATWORK/manifest-v2"