Explicit teaching shipped with C1 (the Spam / Not spam buttons). This adds the
two halves that need no buttons at all.
Passive learning (§C.5). Moving mail into Spam says "this is spam" more honestly
than any button, because the user is doing it for their own reasons rather than
to train a filter; moving it back out says the opposite. Only those two
transitions teach — Spam to Trash is someone tidying up, Inbox to Archive says
nothing about spam. Teaching the same message the same thing twice is skipped,
since users move things back and forth, but reversing a lesson is allowed. If
the engine is down nothing is recorded, so a later move can teach it again.
This needed the transition, not just the destination: `message.changed` carried
only where a message landed, so `move_job` now also publishes
previous_mailbox_id and both special_use values.
Provider bootstrap (§C.5). An account going live enqueues a job that teaches
Bayes from mail the user already owns — up to 500 messages from the provider's
Junk folder as spam, and up to 500 Inbox messages older than 30 days as ham.
This is the constructive use of the history that classification deliberately
refuses to act on: the same backfill, read instead of judged. It sleeps between
messages so live classification never queues behind training, stops rather than
grinding if the engine disappears, and marks the account done in KV so it runs
once.
That needed `PluginMessages.sample`: the SDK could fetch a message by id but not
enumerate a folder. Ids only — bodies are pulled one at a time, so a 500-message
sample never materialises 500 bodies.
Verified against the running engine: teaching moved its `learned` counter 0 -> 2.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018SqJYt7jD7oUEkoY68sBZo
message.new fires for imported history as well as new arrivals, and EventBus
awaits hooks inline. As written, adding an account would have scored every
message in its backfill — filing years-old inbox mail into Junk on the server,
and adding ~70 ms of serialized HTTP per message to the backfill loop (twelve
minutes on a 10k-message account). It is also when the filter is least able to
judge, because Bayes has learned nothing yet.
The envelope now carries account_state, and the hook returns unless it is
"live". The backfill window belongs to the C2 learning bootstrap (§C.5), which
reads that history as training data instead of acting on it.
Tests pin the gating: backfilling accounts are not even read from the blob
store, non-inbox folders are ignored, a live inbox arrival still classifies, and
an unreachable engine still delivers.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018SqJYt7jD7oUEkoY68sBZo
Three additive surfaces the antispam plugin (M7, Appendix C) needs and the
frozen §11 API did not have. None of them changes a core mail path.
`actions.move_to_special` moves ONE message into an account's folder by
special-use, wrapping the same optimistic-update + move_job path the core
endpoint uses. The existing `trash` moves a whole thread, which is right for
Delete and wrong for a spam verdict: one spam reply must not drag a legitimate
conversation into Junk. It returns False rather than raising when the account
has no such folder — an account whose server lacks a Junk folder must still
receive its mail.
`messages.raw` returns the stored RFC822 bytes. Spam scoring needs headers and
body; the SDK could only hand out one header at a time.
`envelope` now also carries account_id and mailbox_special_use. Appendix C must
classify only inbox arrivals, and neither the message.new payload nor the
envelope previously said where a message landed.
docs/open-questions.md gains OQ-6 (why these are additive rather than a spec
violation of "zero core diffs") and OQ-7 (why the engine is reached with plain
httpx instead of the SSRF-guarded sdk.http).
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_018SqJYt7jD7oUEkoY68sBZo
Gmail sync (SPEC: Gmail sync):
- Provider fetches X-GM-LABELS when X-GM-EXT-1 is advertised; parses custom labels
(system labels dropped, mUTF-7 decoded) onto RemoteMessage.
- Engine syncs only system folders + All Mail for Gmail, skips every label folder,
and dedupes the All Mail pass by Message-ID — kills the INBOX/All-Mail duplication.
- Custom labels ride the message.new event; the labels plugin subscribes and maps
them to chips as "<name> [Gmail]" (per-message + thread aggregate + known set),
so labels fill in gradually with the sync window. Non-ASCII labels decode right.
Labels self-healing (message-driven reconcile):
- Per-message keys are the source of truth; t: aggregates and _labels are derived.
Reconcile drops labels of deleted messages, regroups survivors by their current
thread (fixing JWZ thread merges), rewrites changed t: keys, and rebuilds _labels
from live messages — keeping user labels, dropping orphaned Gmail ones.
- New SDK reads: sdk.threads.existing / sdk.messages.threads_of. Sidebar refreshes
live via a plugin.labels.changed WS event on new labels.
Account deletion: prune threads left empty after the cascade (they were user-scoped,
so only messages cascaded) + de-list the account from surviving cross-account threads.
Sync SPAM: opt-in per-account sync_spam (migration 0015, default off) — the engine
skips the Junk folder in backfill and poll unless enabled; wizard checkbox added.
Robustness / UX:
- WebSocket auto-reconnects with backoff and resyncs on (re)connect, and account
add/delete invalidates threads/labels — no more hard refresh for live updates.
- Backfill banner + sidebar ring also require done < total, so a completed account
can't leave a stuck "X of X".
- Scrub lone UTF-16 surrogates from parsed headers/bodies so bad messages no longer
crash the backfill on Postgres text/JSONB.
- Ultimo signet logo in the Mail/Calendar/Contacts sidebar header (and at the top of
the collapsed rail), via the shared plugin-ui.css + MailLayout.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01U72FZiF1W1pb8XxyWBTREA
lukasandClaude Opus 5