audit: decisions of 2026-10-12 written into the plan

Stage 0 then 1+2 agreed; H4 folds logins; H3 measured without a screen
(delivery_log_days, decision rule, no card); L5 validated at save; L12 the
send log gets domain_id; L8 the reply names no other domain; L9 local
comment and test only; stage 5 taken whole. Migrations renumbered
0007-0009.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

docs/audit-2026-10-12.md

@@ -552,30 +552,54 @@ says so; the order is my recommendation by harm per hour.
    does; `reconcile` already treats an error as no information. Test: a 5 MiB
    fake `postqueue -p`. — **Opus**
 2. **Reconcile only closes `queued` rows** (`AND status = 'queued'` in the
-   reconcile path's update) and **only while the tailer is reading**: `follow`
+   reconcile path's update) and marks what it closed: `reason = 'reconcile'`
-   publishes "healthy" (last successful drain within N ticks) and
+   on the row (the 0006 column; a new value, no schema change). — **Opus**
-   `reconcileLoop` skips a sweep when it is not. Test: reopen fails → no rows
+3. **Measure before deciding the sweep's fate** (owner, 2026-10-12: the
-   closed. — **Opus**
+   keep / remove / `lost` question needs real data, and today the only
-3. **Rotation window**: after switching inodes keep draining the old one until
+   record of a sweep is a stdout line that dies with the container). A table
-   it has not grown for one full tick *and* the new file has received a line,
+   `delivery_log_days` (migration `0007`, DDL only; a row per UTC day, kept a
-   or for a bounded number of ticks. Test on the stand with `postfix reload`
+   year, the shape of `sender_check_days`) with four counters:
-   timing. — **Opus**
+   `reconcile_closed` (rows the sweep closed); `reconcile_contradicted` (a
-4. **Anchor `deliveryRe`** at the head of the line like its two siblings;
+   `reason = 'reconcile'` row the tailer later rewrote from a `status=` line —
+   the log wins, and the rewrite is counted); `tailer_blind_seconds`
+   (`follow()` unable to read: reopen failing, file gone — the sender check's
+   downtime mechanics); `queued_over_lifetime` (rows still `queued` past
+   `maximal_queue_lifetime` at the day's end — what removing the sweep would
+   leave visible). **No screen for this** (owner, 2026-10-12: nothing is
+   drawn that ordinary operation will not need): the table is read off a
+   copy of the database when the month is up. Whether a durable *Delivery
+   log* card (tailer blindness, stuck rows — the two counters that outlive
+   the decision) earns a place on Health is a separate question, asked only
+   if the data shows those numbers are ever non-zero.
+   Decision rule, written now: `closed = 0` for a month → remove the sweep;
+   `closed > 0, contradicted = 0` → keep; `contradicted > 0` → each case
+   explained by `tailer_blind` means fix the tailer and keep the sweep,
+   otherwise `lost` or remove. Decide after 30 days of data, or sooner if
+   `contradicted` is ever non-zero. — **Opus**
+4. **Deferred until the data is in:** pausing the sweep while the tailer is
+   not reading (publish "healthy" from `follow`, skip the sweep otherwise),
+   and the rotation window (keep draining the old inode until it has not
+   grown for a tick and the new file has a line). Both are fixes to the
+   tailer's blindness and stay on the list whatever the sweep's fate; their
+   priority follows `tailer_blind_seconds`.
+5. **Anchor `deliveryRe`** at the head of the line like its two siblings;
    test with the probed HELO and quoted-sender lines. — **Opus**
-5. **TLS re-copy** (H2): `postfix-cert-reload.sh` compares the mounted files
+6. **TLS re-copy** (H2): `postfix-cert-reload.sh` compares the mounted files
    to the copy and re-copies before `postfix reload` (atomically, same modes);
    e2e replaces the cert in the stand and reads the new serial off 465.
    — **Opus**
-6. **Deferred counter** (M13): on the stand, `kill -STOP` the panel mid-message
+7. **Deferred counter** (M13): on the stand, `kill -STOP` the panel mid-message
    and record which program logs the warning; widen the matcher accordingly
    and replace the composed fixture with the seen lines. — **Opus**
-7. Docs: architecture § Log tailer, § Queue reconcile, § Sender check "what it
+8. Docs: architecture § Log tailer, § Queue reconcile (the measurement and
-   costs"; guide § Delivery log. — **Sonnet**
+   the rule), § Sender check "what it costs"; guide § Delivery log, § Health.
+   — **Sonnet**
 
-### Stage 2 — the tenant boundary and the registry (H4, M7, M9, M10) → in `2.5.0` (migration `0007`)
+### Stage 2 — the tenant boundary and the registry (H4, M7, M9, M10) → in `2.5.0` (migration `0008`)
 
-1. **Logins lower-cased** at validation and `applications.login COLLATE
+1. **Logins lower-cased** at validation (owner, 2026-10-12: fold, do not
-   NOCASE` (migration rebuilds the unique index). Before the migration a check
+   refuse) and `applications.login COLLATE NOCASE` (migration `0008` rebuilds
+   the unique index). Before the migration a check
    that no two existing logins differ only in case (there is one
    installation; look). Tests: `Ops`/`ops` refused; the map written lower-
    case. Then verify Postfix's compare on the stand once (`MAIL FROM` as the
@@ -588,16 +612,37 @@ says so; the order is my recommendation by harm per hour.
 4. **Import validates like the form**: route `DMARCRua`, IPs, limit mode,
    window and count through the same parsers; refuse the file with the
    validator's text. Tests per field. — **Opus**
-5. Docs: security.md § Roles (the guard), guide § Domain export. — **Sonnet**
+5. **Send-log rows belong to a domain by id** (L12; owner, 2026-10-12).
+   `send_log.domain_id INTEGER REFERENCES domains(id) ON DELETE SET NULL` in
+   migration `0008`, backfilled once by name against the registry as it is at
+   migration time (rows of a domain already gone stay NULL); written by the
+   journal milter and by the sender check's refusal rows at insert, looked up
+   from the login's domain; the `domain` name column stays for display and
+   for the global view. A domain user's log, stats, auto rate limit and the
+   notifier's refusal count select by id; a NULL row is nobody's but the
+   global role's. A domain deleted and re-added under its old name therefore
+   starts with an empty log. The inbound journal (`inbound_spam_log`, the
+   same accepted risk in security.md:419) gets the same treatment in the same
+   migration — my proposal, so that the two logs do not follow two rules.
+   Tests: delete-and-re-add shows nothing to the new assignee; the old rows
+   are still in the global log. — **Opus**
+6. Docs: security.md § Roles (the guard) and the struck accepted risk, guide
+   § Domain export, schema-migrations.md. — **Sonnet**
+7. **The 550 reply names no other domain** (L8): `sender.go:334` stops at
+   "it names another sender of this server"; the refused message's page
+   still shows the `From:` line as written (`header_from`), so the operator
+   loses nothing. Test pins the reply text; e2e's sender-check negatives
+   assert the reply does not contain the other tenant's name. — **Opus**
 
 ### Stage 3 — persistence and restore (M1–M5, L1–L4, L7) → `2.5.1`
 
 1. **Consume the manifest after the resync succeeds** (`CheckRestore` reports;
    `run` deletes after `resyncAfterRestore` returns nil). Test: Resync fails →
    manifest kept → second boot resyncs. — **Opus**
-2. **Indexes**: `idx_send_log_domain_created_at (domain, created_at)`;
+2. **Indexes**: `idx_send_log_domain_created_at (domain_id, created_at)` —
+   on the id column stage 2 adds, replacing `idx_send_log_domain`;
    partial `idx_send_log_queued (created_at) WHERE status = 'queued'`;
-   consider `(domain, id)` for the scoped page. Migration `0008` DDL only;
+   consider `(domain, id)` for the scoped page. Migration `0009` DDL only;
    `TestLimitRefusalsSince`-style EXPLAIN tests against the **real** query
    strings. — **Opus**
 3. **CLI backup**: open the source `mode=ro`, refuse a missing file; password
@@ -629,6 +674,12 @@ says so; the order is my recommendation by harm per hour.
    the flash and the two comments say what it covers. — **Sonnet**
 5. `domain.Service.Delete`: say the domain is deleted when a later step fails,
    and run `apps.Resync()` anyway. — **Opus**
+6. **Upstream TLS mode validated at save** (L5): saving an inbound domain
+   whose `[host]:port` is already used by another domain with a different
+   TLS mode is refused with the validator's text naming that domain and its
+   mode; `TestRenderInboundMaps` gets the two-routes-one-next-hop case and the
+   map is asserted to hold one key. Guide § Inbound domains says a next hop
+   has one TLS mode. — **Opus**
 
 ### Stage 5 — image and Postfix hardening (M17–M19, M21, L30–L32) → `2.6.0`
 
@@ -676,14 +727,17 @@ says so; the order is my recommendation by harm per hour.
 
 ### Not planned, said out loud
 
-- L5 (`tls_policy` per next-hop): the right fix is a model change (TLS mode
+- L5 (`tls_policy` per next-hop): the full fix is a model change (TLS mode per
-  per upstream, not per domain) and a form change — a design item, not a
+  upstream, not per domain) — a design item. Decided (owner, 2026-10-12):
-  patch. Until then the guide should say two domains to one upstream share
+  validate instead — stage 4, item 6.
-  the first one's TLS mode.
+- L12 (rows by domain name): decided (owner, 2026-10-12): the send log gets
-- L12 (rows by domain name): accept it for outbound as security.md already
+  the domain's id — stage 2, item 5.
-  does for inbound, and write it there.
+- L8 (the signing-name oracle): decided (owner, 2026-10-12): do it — stage 2,
-- L8 (the signing-name oracle): the names are in DNS; shortening the reply
+  item 7. (The names are not public: rDNS gives the hostname only, so the
-  costs a less useful error. Leave, note in security.md § Accepted risks.
+  reply told one tenant which other domains are hosted here.)
-- L9 (go-milter header accumulation): a one-line comment on `Headers` and a
+- L9 (go-milter header accumulation): decided (owner, 2026-10-12): locally
-  test with a polluted `h`; the library behaviour itself is upstream's.
+  only — a comment on `senderSession.Headers` saying why `h` is ignored, and
+  `TestSenderCheckResetsBetweenMessages` passing an `h` that still holds the
+  first message's `From:` (stage 2, with item 7). The library itself stays
+  as vendored; an upstream patch is a separate, unhurried matter.
 - The Windows-bound tests: not worth platform guards; CI is Linux.