Read a web page when he names one, and watch a few on a timer (#259)

The network fallback behind the local sources, off unless configured.

internal/crawl is pure: a stdlib robots.txt parser (group specificity,
wildcards, Crawl-delay, cached per host), HTML-to-plaintext extraction, and a
watcher that notes a watched page only when its text changed. It has no store
access and no net/http; cmd/mavend/crawls.go is the impure half.

Every limit is code and tested: the guarded fetcher from #258 enforces the host
allowlist/denylist, refuses private addresses in the dialer Control hook (so DNS
rebinding and each redirect hop are covered), caps size and redirects, times out,
and spaces requests per host. A robots.txt Disallow is refused with no override.

On demand, reading is a query source placed last in the chain, after his memory,
his notes, and the local Kiwix ZIMs once those are wired: no URL in the
utterance means no fetch, and only the URL ever leaves the box. Scheduled
watches write notes and announce nothing.

The vendored tree has no x/net/html, goquery or temoto/robotstxt, so the parsers
are stdlib. No new dependency.

.gitignore

@@ -7,6 +7,7 @@
 /mavpoll
 /mavcaldav
 /mavwaked
+/mavmaild
 
 # Certs (private keys, don't commit)
 certs/
@@ -34,6 +35,10 @@ deps
 deploy/db_key.env
 # Deploy secret (telegram bot token + chat id) — never commit
 deploy/telegram.env
+# zenmoney API token, read by mavpoll (never in argv, never committed)
+deploy/zenmoney.token
+# IMAP password, read by mavmaild (never in argv, never committed)
+deploy/imap.password
 
 # Temp files
 /tmp/

CLAUDE.md

@@ -34,7 +34,7 @@ CGO daemons (`mavend`, `mavsttd`, `mavttsd`, `mavenclient`) need the vendored to
 and libs wired through the Makefile — **do not** call `go build` on them bare, use `make`:
 
 ```sh
-make build          # all 8 binaries
+make build          # all 9 binaries
 make build-web      # single daemon (pure-Go ones: web/waked/poll/caldav build without CGO)
 make test           # go test -race across ./internal/... ./cmd/... with CGO env set
 ```
@@ -62,6 +62,7 @@ Pure-Go packages (`router`, `memory`, `mavweb`, …) run under a plain `go test
 | `mavenclient` | Voice loop client (mic → stt → core → tts). |
 | `mavpoll` | Telegram long-poll reach. |
 | `mavcaldav` | CalDAV calendar sync. |
+| `mavmaild` | Mail reader (IMAP, read-only). Holds the IMAP password; core never sees it. |
 
 Daemons are wired socket-to-socket, not linked. `internal/ipc` is the client/server wire
 protocol; the config in `deploy/mavend.json` (with `${VAR}` env expansion from gitignored

Dockerfile

@@ -51,7 +51,8 @@ RUN go build -o /out/mavend    ./cmd/mavend    && \
     go build -o /out/mavttsd   ./cmd/mavttsd   && \
     go build -o /out/mavweb    ./cmd/mavweb    && \
     go build -o /out/mavpoll   ./cmd/mavpoll   && \
-    go build -o /out/mavcaldav ./cmd/mavcaldav
+    go build -o /out/mavcaldav ./cmd/mavcaldav && \
+    go build -o /out/mavmaild  ./cmd/mavmaild
 
 # llama.cpp Vulkan build — the phraser/router LFM engine (llama-server). Built
 # from source (not a prebuilt vendored blob) so the binary's glibc/GLIBCXX match

Makefile

@@ -20,7 +20,7 @@ PIPER_ESPEAK  := $(shell pwd)/deps/piper/espeak-ng-data
 
 all: build
 
-build: build-stt build-tts build-daemon build-client build-waked build-web build-poll build-caldav
+build: build-stt build-tts build-daemon build-client build-waked build-web build-poll build-caldav build-mail
 
 build-stt:
 	CGO_CFLAGS="$(CGO_CFLAGS)" CGO_LDFLAGS="$(CGO_LDFLAGS)" LD_LIBRARY_PATH="$(shell pwd)/deps/lib" \
@@ -50,6 +50,9 @@ build-poll:
 build-caldav:
 	$(GO) build $(GOFLAGS) -o mavcaldav ./cmd/mavcaldav/
 
+build-mail:
+	$(GO) build $(GOFLAGS) -o mavmaild ./cmd/mavmaild/
+
 run-web: build-web
 	./mavweb -addr :9200 -voice 127.0.0.1:9100
 
@@ -185,4 +188,4 @@ download-embedder:
 	@echo '  sudo cp onnxruntime-linux-x64-1.15.1/lib/libonnxruntime.so* /usr/local/lib/'
 
 clean:
-	rm -f mavend mavenclient mavsttd mavttsd mavweb mavpoll mavcaldav mavwaked
+	rm -f mavend mavenclient mavsttd mavttsd mavweb mavpoll mavcaldav mavwaked mavmaild

cmd/mavend/actions_money.go

@@ -0,0 +1,78 @@
+package main
+
+import (
+	"context"
+	"log"
+
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/zenmoney"
+)
+
+// Money questions (Vikunja #125).
+//
+// This is the whole read side: mavpoll holds the zenmoney token and writes
+// facts(kind=env, source=poll:zenmoney); core reads them back when he asks.
+// Core never sees the token, never calls zenmoney, and has no rule on these
+// keys — a total is never a reason for Maven to speak first. Maven is not a
+// nag, least of all about his money.
+//
+// Nothing here can reach the external search capability: the figures are read
+// from the store and rendered locally, and his financial data is never search
+// input.
+
+// queryMoney — "сколько я потратил сегодня?", "покажи мои траты".
+//
+// Answers only from the latest fact the poller wrote. Three honest outcomes and
+// no fourth: the figure, "the fact is old and here is its date", or "money
+// tracking is not connected". It never computes, estimates or rounds a total of
+// its own — an invented number about his money is the worst thing this could do.
+func (h *reactiveHandler) queryMoney(ctx context.Context, t *queryTurn) (string, bool) {
+	window, ok := router.ParseMoneyQuery(t.dec.Utterance)
+	if !ok {
+		return "", false
+	}
+	key, phrase := zenmoney.KeySpentMonth, "в этом месяце"
+	if window == router.MoneyToday {
+		key, phrase = zenmoney.KeySpentToday, "сегодня"
+	}
+	fact, err := h.api.LatestFactBySource(ctx, key, zenmoney.Source)
+	if err != nil {
+		// No fact at all is the normal state when the capability is off. Claim
+		// the turn anyway: falling through to recall would answer a question
+		// about money with whatever note happens to be nearest.
+		if !isNoFactErr(err) {
+			log.Printf("voice: money fact: %v", err)
+		}
+		return "я не отслеживаю траты — не подключено.", true
+	}
+	val, err := zenmoney.ParseFactValue(fact.Value)
+	if err != nil {
+		log.Printf("voice: money fact: decode: %v", err)
+		return "не получилось прочитать траты.", true
+	}
+	reply := val.FormatRU(phrase)
+	if reply == "" {
+		return "по тратам пока нечего сказать.", true
+	}
+	// A stale fact is reported as stale rather than spoken as today's number.
+	if h.now().Sub(fact.Ts) > zenmoney.StaleAfter {
+		return "данные от " + fact.Ts.Local().Format("02.01") + ": " + reply, true
+	}
+	return reply, true
+}
+
+// isNoFactErr — ErrNoFact survives the wire wrapped, so unwrap for it.
+func isNoFactErr(err error) bool {
+	for e := err; e != nil; {
+		if e == ipc.ErrNoFact {
+			return true
+		}
+		u, ok := e.(interface{ Unwrap() error })
+		if !ok {
+			return false
+		}
+		e = u.Unwrap()
+	}
+	return false
+}

cmd/mavend/actions_money_test.go

@@ -0,0 +1,139 @@
+package main
+
+import (
+	"context"
+	"strings"
+	"testing"
+	"time"
+
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/zenmoney"
+)
+
+// moneyAPI answers only LatestFactBySource; everything else is unimplemented,
+// which is the assertion that answering a money question costs no model call
+// and reaches no network.
+type moneyAPI struct {
+	ipc.UnimplementedCoreAPI
+
+	fact    ipc.Fact
+	err     error
+	gotKey  string
+	gotSrc  string
+	callCnt int
+}
+
+func (a *moneyAPI) LatestFactBySource(_ context.Context, key, source string) (ipc.Fact, error) {
+	a.gotKey, a.gotSrc = key, source
+	a.callCnt++
+	return a.fact, a.err
+}
+
+func moneyNow() time.Time { return time.Date(2026, 8, 15, 20, 0, 0, 0, time.UTC) }
+
+func moneyFact(ts time.Time, val string) ipc.Fact {
+	return ipc.Fact{Kind: "env", Key: zenmoney.KeySpentMonth, Value: val, Source: zenmoney.Source, Ts: ts}
+}
+
+func TestQueryMoneyAnswersFromTheFact(t *testing.T) {
+	api := &moneyAPI{fact: moneyFact(moneyNow(), `{"spent":[{"currency":"RUB","amount":1749.5}],"count":3}`)}
+	h := &reactiveHandler{api: api, now: moneyNow}
+	reply, ok := h.queryMoney(context.Background(), &queryTurn{
+		dec: router.Decision{Utterance: "сколько я потратил в этом месяце?"},
+	})
+	if !ok {
+		t.Fatal("the money source must claim a money question")
+	}
+	if api.gotKey != zenmoney.KeySpentMonth || api.gotSrc != zenmoney.Source {
+		t.Errorf("read %q/%q, want the month key from the poller's source", api.gotKey, api.gotSrc)
+	}
+	if !strings.Contains(reply, "1749.5") {
+		t.Errorf("reply = %q, want the exact figure", reply)
+	}
+	if !strings.Contains(reply, "в этом месяце") {
+		t.Errorf("reply = %q, want the window named", reply)
+	}
+}
+
+func TestQueryMoneyPicksTodaysKey(t *testing.T) {
+	api := &moneyAPI{fact: moneyFact(moneyNow(), `{"spent":[{"currency":"RUB","amount":250}],"count":1}`)}
+	h := &reactiveHandler{api: api, now: moneyNow}
+	if _, ok := h.queryMoney(context.Background(), &queryTurn{
+		dec: router.Decision{Utterance: "сколько я потратил сегодня?"},
+	}); !ok {
+		t.Fatal("expected the source to claim it")
+	}
+	if api.gotKey != zenmoney.KeySpentToday {
+		t.Errorf("key = %q, want today's", api.gotKey)
+	}
+}
+
+// The capability is off unless configured, and then there is no fact. She says
+// so instead of letting the recall pass answer a money question from a note.
+func TestQueryMoneySaysNotConnected(t *testing.T) {
+	h := &reactiveHandler{api: &moneyAPI{err: ipc.ErrNoFact}, now: moneyNow}
+	reply, ok := h.queryMoney(context.Background(), &queryTurn{
+		dec: router.Decision{Utterance: "сколько я потратил?"},
+	})
+	if !ok {
+		t.Fatal("expected the source to claim it")
+	}
+	if !strings.Contains(reply, "не подключено") {
+		t.Errorf("reply = %q, want an honest 'not connected'", reply)
+	}
+	// No number of any kind in that answer.
+	for _, d := range []string{"0", "1", "2", "3", "4", "5", "6", "7", "8", "9"} {
+		if strings.Contains(reply, d) {
+			t.Errorf("reply %q contains a digit — nothing was read, so there is no figure", reply)
+		}
+	}
+}
+
+// A fact older than the staleness bound is dated rather than spoken as if it
+// were current: the poller can be down, and last week's total presented as
+// today's is a lie by omission.
+func TestQueryMoneyDatesAStaleFact(t *testing.T) {
+	old := moneyNow().Add(-72 * time.Hour)
+	api := &moneyAPI{fact: moneyFact(old, `{"spent":[{"currency":"RUB","amount":100}],"count":1}`)}
+	h := &reactiveHandler{api: api, now: moneyNow}
+	reply, _ := h.queryMoney(context.Background(), &queryTurn{
+		dec: router.Decision{Utterance: "сколько я потратил?"},
+	})
+	if !strings.Contains(reply, "данные от") {
+		t.Errorf("reply = %q, want the stale fact dated", reply)
+	}
+}
+
+func TestQueryMoneyPassesOtherQuestions(t *testing.T) {
+	api := &moneyAPI{}
+	h := &reactiveHandler{api: api, now: moneyNow}
+	for _, u := range []string{"какая погода?", "я потратил весь день на это", "какие у меня задачи?"} {
+		if _, ok := h.queryMoney(context.Background(), &queryTurn{dec: router.Decision{Utterance: u}}); ok {
+			t.Errorf("the money source claimed %q", u)
+		}
+	}
+	if api.callCnt != 0 {
+		t.Error("a non-money question must not read the money facts")
+	}
+}
+
+// Money must be answered before the recall sources, or a question about
+// spending gets answered by the nearest note.
+func TestQuerySourcesOrderMoneyBeforeRecall(t *testing.T) {
+	moneyAt, notesAt := -1, -1
+	for i, src := range querySources {
+		switch src.name {
+		case "money":
+			moneyAt = i
+		case "notes":
+			notesAt = i
+		}
+	}
+	if moneyAt < 0 || notesAt < 0 {
+		t.Fatalf("sources missing: money=%d notes=%d", moneyAt, notesAt)
+	}
+	if moneyAt > notesAt {
+		t.Errorf("money source at %d, after notes at %d", moneyAt, notesAt)
+	}
+}

cmd/mavend/actions_query.go

@@ -8,10 +8,12 @@ import (
 	"strings"
 	"time"
 
+	"github.com/kami/maven/internal/crawl"
 	"github.com/kami/maven/internal/ipc"
 	"github.com/kami/maven/internal/memory"
 	"github.com/kami/maven/internal/morning"
 	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/rss"
 	"github.com/kami/maven/internal/weather"
 )
 
@@ -62,11 +64,28 @@ var querySources = []querySource{
 	// matcher requires a task noun or an explicit "что … сделать", so a
 	// date-bearing question still reaches the calendar.
 	{"tasks", (*reactiveHandler).queryTasks},
+	// Before the recall sources too: "сколько я потратил?" is a question about
+	// the money facts the poller wrote, and the notes pass would otherwise
+	// answer it from whatever he once said about spending. Its matcher needs a
+	// money noun plus an actual ask, so "я потратил весь день" is untouched.
+	{"money", (*reactiveHandler).queryMoney},
+	// Before the recall sources and before general knowledge: "что нового?" is
+	// a question about the feeds she reads, and general knowledge would answer
+	// it by inventing news. Its matcher needs a feed noun plus an ask, so
+	// "у меня новая лента в инстаграме" is untouched.
+	{"feeds", (*reactiveHandler).queryFeeds},
 	{"calendar", (*reactiveHandler).queryCalendar},
 	{"weather", (*reactiveHandler).queryWeather},
 	{"embed", (*reactiveHandler).queryEmbed},
 	{"memory", (*reactiveHandler).queryMemory},
 	{"notes", (*reactiveHandler).queryNotes},
+	// LAST before the model answers from memory, and that position is the whole
+	// design (Vikunja #259): local sources first. The model, his own notes and
+	// facts, and — once internal/kiwix is wired into this chain — the offline
+	// ZIMs all get their turn before anything touches the network. This source
+	// only claims a turn where he named a URL out loud, so it never competes
+	// with a local answer.
+	{"web", (*reactiveHandler).queryWeb},
 	{"general-knowledge", (*reactiveHandler).queryGeneral},
 }
 
@@ -175,6 +194,63 @@ func (h *reactiveHandler) queryHabits(ctx context.Context, t *queryTurn) (string
 	return profile.FormatOverallRU(), true
 }
 
+// feedNoteWindow — how many recent notes are scanned for feed items, and
+// feedReadOut — how many headlines she actually reads back. She summarises the
+// top of the pile, she does not recite a river.
+const (
+	feedNoteWindow = 200
+	feedReadOut    = 3
+)
+
+// queryFeeds — "что нового в лентах?", "что нового по технологиям?"
+// (Vikunja #258).
+//
+// This is the ONLY way a feed item reaches him. The poller writes notes and
+// never speaks; asking is the trigger. If that ever changes, the thing that
+// changed is "Maven is not a nag", not a detail of this file.
+func (h *reactiveHandler) queryFeeds(ctx context.Context, t *queryTurn) (string, bool) {
+	q, ok := router.ParseFeedQuery(t.dec.Utterance)
+	if !ok {
+		return "", false
+	}
+	if !h.feedsOn {
+		// Claim the turn rather than fall through: "не читаю ленты" is true, and
+		// letting general knowledge answer "что нового?" would be an invented
+		// news bulletin.
+		return "я пока не читаю ленты — они не настроены.", true
+	}
+	notes, err := h.api.RecentNotes(ctx, feedNoteWindow)
+	if err != nil {
+		log.Printf("voice: feeds: recent notes: %v", err)
+		return "не получилось посмотреть ленты.", true
+	}
+	var picked []string
+	for _, n := range notes {
+		if !strings.HasPrefix(n.Source, rss.SourcePrefix) {
+			continue
+		}
+		if !router.CategoryMatches(n.Text, q.Category) {
+			continue
+		}
+		// The note carries title, summary and link; she reads the title.
+		title := n.Text
+		if i := strings.IndexByte(title, '\n'); i > 0 {
+			title = title[:i]
+		}
+		picked = append(picked, strings.TrimSpace(title))
+		if len(picked) == feedReadOut {
+			break
+		}
+	}
+	if len(picked) == 0 {
+		if q.Category != "" {
+			return "по этой теме в лентах пока ничего.", true
+		}
+		return "в лентах пока ничего нового.", true
+	}
+	return "вот что нового: " + strings.Join(picked, "; "), true
+}
+
 // queryCalendar — "что у меня сегодня?", "планы на завтра?"
 // h.now(), not time.Now(): the handler's clock is the injected one, so this
 // source can be tested at a fixed time like the rest.
@@ -303,6 +379,57 @@ func (h *reactiveHandler) queryNotes(ctx context.Context, t *queryTurn) (string,
 	return reply, true
 }
 
+// webPageContextRunes — how much of a fetched page is handed to the phraser.
+// Less than the crawler keeps: the rest of the 4096-token window belongs to the
+// prompt, the persona block and the reply.
+const webPageContextRunes = 1500
+
+// queryWeb — "посмотри https://example.org/x — что там?" (Vikunja #259).
+//
+// It claims a turn ONLY when he named a URL, which is what keeps a fallback from
+// becoming a habit: no URL, no fetch, and the model answers from what is local.
+// What leaves the box is the URL and nothing else — no note, no fact, no history
+// travels with it.
+func (h *reactiveHandler) queryWeb(ctx context.Context, t *queryTurn) (string, bool) {
+	link, ok := router.FirstURL(t.dec.Utterance)
+	if !ok {
+		return "", false
+	}
+	if h.crawler == nil {
+		// Claim rather than fall through: he asked about a specific page, and
+		// letting the model answer from the URL's spelling alone is how a small
+		// model invents a page's contents.
+		return "я не читаю страницы — это не настроено.", true
+	}
+	ctxFetch, cancel := context.WithTimeout(ctx, 30*time.Second)
+	defer cancel()
+	page, err := h.crawler.Page(ctxFetch, link)
+	if err != nil {
+		if errors.Is(err, crawl.ErrRobots) {
+			return "эта страница закрыта для чтения — robots.txt не разрешает.", true
+		}
+		log.Printf("voice: web: %v", err)
+		return "не получилось прочитать страницу.", true
+	}
+	if page.Text == "" {
+		return "страница открылась, но читать там нечего.", true
+	}
+	// The page is handed to the phraser the same way a note is: as context for
+	// the question he actually asked. She answers the question, she does not
+	// recite the page.
+	snippet := page.Title + "\n" + crawl.TrimRunes(page.Text, webPageContextRunes)
+	reply, perr := h.phraser.PhraseQuery(ctx, t.dec.Utterance, []string{snippet})
+	if perr != nil {
+		log.Printf("voice: web: phrase: %v", perr)
+	}
+	if reply == "" {
+		// No phraser (or it failed): read back the top of the page rather than
+		// pretend the fetch did not happen.
+		return "вот что на странице: " + crawl.TrimRunes(page.Text, 300), true
+	}
+	return reply, true
+}
+
 // queryGeneral — general knowledge from the phraser, the last source before
 // giving up. It always claims: either the model answers or Maven says she
 // doesn't know.

cmd/mavend/actions_task.go

@@ -3,11 +3,11 @@ package main
 import (
 	"context"
 	"log"
-	"strings"
 
 	"github.com/kami/maven/internal/ipc"
 	"github.com/kami/maven/internal/router"
 	"github.com/kami/maven/internal/store"
+	"github.com/kami/maven/internal/tasks"
 )
 
 // Task capture on the voice/chat path (Vikunja #130).
@@ -27,14 +27,15 @@ import (
 // captureTaskFromNote claims the turn when the utterance explicitly files a
 // task, returning the reply. ("", false) hands the turn back to the note path.
 func (h *reactiveHandler) captureTaskFromNote(ctx context.Context, dec router.Decision) (string, bool) {
-	text, ok := router.ParseTaskCapture(dec.Utterance)
+	cap, ok := router.ParseTaskCapture(dec.Utterance)
 	if !ok {
 		return "", false
 	}
 	resp, err := h.api.CaptureTask(ctx, ipc.CaptureTaskReq{
-		Text:   text,
+		Text:   cap.Text,
 		Source: "tap:voice",
 		Status: store.TaskOpen, // he stated it himself — not a candidate
+		Weight: cap.Weight,     // 0 unless he said "срочно" / "важно"
 		Ts:     h.now(),
 	})
 	if err != nil {
@@ -44,56 +45,39 @@ func (h *reactiveHandler) captureTaskFromNote(ctx context.Context, dec router.De
 	if !resp.Created {
 		return "это уже в списке.", true
 	}
-	return "записала: " + text, true
+	return "записала: " + cap.Text, true
 }
 
 // queryTasks — "какие у меня задачи?", "что мне нужно сделать?".
 //
-// Reads the live set and recites it. Newest first, which is the order the store
-// returns: this source has no opinion about which task matters more, and
-// pretending otherwise would be a guess. Ranking is Vikunja #129.
+// Reads the live set and recites it in priority order (Vikunja #129). The order
+// is computed by internal/tasks from what he told her — deadlines, the urgency
+// he stated, how long a task has been sitting — never asked of the model. The
+// rendering is the package's too, so the spoken list and the /tasks page can
+// never disagree about what comes first.
 func (h *reactiveHandler) queryTasks(ctx context.Context, t *queryTurn) (string, bool) {
 	if !router.IsTaskListQuery(t.dec.Utterance) {
 		return "", false
 	}
-	tasks, err := h.api.ListTasks(ctx, "live")
+	live, err := h.api.ListTasks(ctx, "live")
 	if err != nil {
 		log.Printf("voice: list tasks: %v", err)
 		return "не получилось посмотреть задачи.", true
 	}
-	return formatTaskListRU(tasks), true
+	return tasks.FormatRU(tasks.Rank(taskItems(live), h.now())), true
 }
 
-// formatTaskListRU renders the live task list the way Maven says it. Candidates
-// are named as candidates — a task she pulled out of his mail is something she
-// suggests, and saying it in the same breath as work he actually stated would
-// put words in his mouth.
-func formatTaskListRU(tasks []ipc.Task) string {
-	var open, cands []string
-	for _, t := range tasks {
-		switch t.Status {
-		case store.TaskCandidate:
-			cands = append(cands, t.Text)
-		default:
-			open = append(open, t.Text)
+// taskItems maps wire rows onto the ranker's input. Written here rather than in
+// internal/tasks so the ranker stays a pure package with no ipc (and therefore
+// no store, and therefore no cgo) dependency — the same posture as
+// internal/morning and internal/memory.
+func taskItems(ts []ipc.Task) []tasks.Item {
+	out := make([]tasks.Item, len(ts))
+	for i, t := range ts {
+		out[i] = tasks.Item{
+			ID: t.ID, Text: t.Text, Status: t.Status,
+			Created: t.CreatedTs, Due: t.Due, Weight: t.Weight,
 		}
 	}
-	if len(open) == 0 && len(cands) == 0 {
-		return "задач нет."
-	}
-	var b strings.Builder
-	if len(open) > 0 {
-		b.WriteString("в списке: ")
-		b.WriteString(strings.Join(open, "; "))
-		b.WriteString(".")
-	}
-	if len(cands) > 0 {
-		if b.Len() > 0 {
-			b.WriteString(" ")
-		}
-		b.WriteString("ещё я нашла, но ты не подтвердил: ")
-		b.WriteString(strings.Join(cands, "; "))
-		b.WriteString(".")
-	}
-	return b.String()
+	return out
 }

cmd/mavend/actions_task_test.go

@@ -141,6 +141,45 @@ func TestQueryTasksRecitesTheLiveList(t *testing.T) {
 	}
 }
 
+// The stated urgency rides through capture as a weight, so the ranker can use
+// it later (Vikunja #129). "срочно" is not part of the task text.
+func TestCaptureTaskCarriesStatedUrgency(t *testing.T) {
+	api := &taskAPI{created: true}
+	h := taskHandler(api)
+	if _, ok := h.captureTaskFromNote(context.Background(), router.Decision{
+		Utterance: "добавь в задачи срочно оплатить интернет",
+	}); !ok {
+		t.Fatal("expected a capture")
+	}
+	got := api.captured[0]
+	if got.Text != "оплатить интернет" {
+		t.Errorf("text = %q, want the urgency word out of the task", got.Text)
+	}
+	if got.Weight == 0 {
+		t.Error("weight = 0 — he said срочно and it was dropped")
+	}
+}
+
+// The recital is ordered by the ranker, not by insertion: a deadline he named
+// comes before undated work.
+func TestQueryTasksRecitesInPriorityOrder(t *testing.T) {
+	due := taskNow()
+	api := &taskAPI{tasks: []ipc.Task{
+		{ID: 1, Text: "купить молоко", Status: "open", CreatedTs: taskNow()},
+		{ID: 2, Text: "оплатить интернет", Status: "open", CreatedTs: taskNow(), Due: &due},
+	}}
+	h := taskHandler(api)
+	reply, _ := h.queryTasks(context.Background(), &queryTurn{
+		dec: router.Decision{Utterance: "какие у меня задачи?"},
+	})
+	if strings.Index(reply, "оплатить интернет") > strings.Index(reply, "купить молоко") {
+		t.Errorf("reply = %q, want the dated task first", reply)
+	}
+	if !strings.Contains(reply, "сегодня") {
+		t.Errorf("reply = %q, want the reason named", reply)
+	}
+}
+
 func TestQueryTasksEmptyList(t *testing.T) {
 	h := taskHandler(&taskAPI{})
 	reply, ok := h.queryTasks(context.Background(), &queryTurn{

cmd/mavend/crawls.go

@@ -0,0 +1,184 @@
+// mavend/crawls.go — the driver for reading web pages (Vikunja #259,
+// docs/plans/14-web-crawler.md). The crawler is pure and lives in
+// internal/crawl; this is the impure half: the guarded fetcher, a ticker for the
+// scheduled watches, and the fact-backed dedup hashes.
+//
+// Two paths, one config block, both off unless configured:
+//
+//   - ON DEMAND — he names a URL out loud and she reads it. That is the
+//     `queryWeb` source in actions_query.go, LAST in the chain: after his
+//     memory, after the notes, and (once Kiwix is wired into the chain) after
+//     the local ZIMs. A local read costs nothing and leaks nothing; a fetch puts
+//     a URL in someone's log, so it goes last.
+//   - SCHEDULED — a watched page is re-read on its interval, and a page whose
+//     text changed is written as a note. It does NOT announce itself. Same rule
+//     as the feed poller: notes, never nudges.
+//
+// Only the URL goes out. Nothing here reads a note, a fact, the persona block or
+// the history, and internal/crawl has no access to the store at all.
+package main
+
+import (
+	"context"
+	"log"
+	"net/url"
+	"time"
+
+	"github.com/kami/maven/internal/config"
+	"github.com/kami/maven/internal/crawl"
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/webfetch"
+)
+
+// newCrawler builds the crawler from the `crawl` block, or returns nil when
+// there is none. Every caller checks for nil, and nil means no page is ever
+// fetched.
+func newCrawler(cfg *config.Config) *crawl.Crawler {
+	if cfg.Crawl == nil {
+		return nil
+	}
+	cc := cfg.Crawl
+
+	hosts := append([]string(nil), cc.AllowHosts...)
+	// A watched page's own host is always reachable; otherwise an allowlist and
+	// a watch list would have to be kept in sync by hand.
+	for _, w := range cc.Watches {
+		if u, err := url.Parse(w.URL); err == nil && u.Hostname() != "" {
+			hosts = append(hosts, u.Hostname())
+		}
+	}
+	// An allowlist plus on-demand is a contradiction worth logging rather than
+	// silently resolving: he asked for arbitrary pages AND for a fixed list.
+	// The allowlist wins, because it is the narrower instruction.
+	if len(hosts) > 0 && cc.OnDemand && len(cc.AllowHosts) > 0 {
+		log.Printf("crawl: allow_hosts is set, so on-demand reading is limited to those hosts")
+	}
+	ua := cc.UserAgent
+	if ua == "" {
+		ua = webfetch.DefaultUserAgent
+	}
+	fetcher := webfetch.New(webfetch.Config{
+		AllowHosts: hosts,
+		DenyHosts:  cc.DenyHosts,
+		Timeout:    time.Duration(cc.Timeout),
+		MaxBytes:   cc.MaxBytes,
+		UserAgent:  ua,
+	})
+	// The user-agent handed to the crawler is the one the fetcher sends: obeying
+	// robots rules written for a different name would be a lie.
+	return crawl.New(&crawlFetcher{f: fetcher}, crawl.Config{
+		UserAgent: ua,
+		MaxRunes:  cc.MaxRunes,
+	})
+}
+
+// onDemandCrawler returns a crawler for the answer path, or nil when on-demand
+// reading is off. The scheduled watches can be on while this is off: reading a
+// fixed list of pages on a timer and reading whatever URL is in an utterance are
+// different permissions, and the config keeps them separate.
+func onDemandCrawler(cfg *config.Config) *crawl.Crawler {
+	if cfg.Crawl == nil || !cfg.Crawl.OnDemand {
+		return nil
+	}
+	return newCrawler(cfg)
+}
+
+// crawlWorker — ticker + watcher for the scheduled half.
+type crawlWorker struct {
+	watcher  *crawl.Watcher
+	interval time.Duration
+}
+
+// crawlTickInterval — how often the worker asks what is due. Per-watch cadence
+// is the watcher's business.
+const crawlTickInterval = 15 * time.Minute
+
+// newCrawlWorker wires the scheduled crawls, or nil when nothing is watched.
+func newCrawlWorker(c *crawl.Crawler, api ipc.CoreAPI, emb router.Embedder, cfg *config.Config) *crawlWorker {
+	if c == nil || cfg.Crawl == nil || len(cfg.Crawl.Watches) == 0 {
+		return nil
+	}
+	watches := make([]crawl.WatchConfig, 0, len(cfg.Crawl.Watches))
+	for _, w := range cfg.Crawl.Watches {
+		watches = append(watches, crawl.WatchConfig{
+			Name:     w.Name,
+			URL:      w.URL,
+			Interval: time.Duration(w.Interval),
+		})
+	}
+	watcher := crawl.NewWatcher(c, watches, api, &factHashes{api: api},
+		crawlEmbedder(emb), time.Duration(cfg.Crawl.Interval))
+	if watcher == nil {
+		log.Printf("crawl: configured but nothing watchable — scheduled crawls disabled")
+		return nil
+	}
+	log.Printf("crawl: watching %d page(s), checking what is due every %s", len(watches), crawlTickInterval)
+	return &crawlWorker{watcher: watcher, interval: crawlTickInterval}
+}
+
+// run checks what is due until ctx is canceled. The first round runs
+// immediately; it writes notes only, so an early round startles nobody.
+func (w *crawlWorker) run(ctx context.Context) {
+	w.watcher.CheckDue(ctx, time.Now())
+	t := time.NewTicker(w.interval)
+	defer t.Stop()
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		case now := <-t.C:
+			w.watcher.CheckDue(ctx, now)
+		}
+	}
+}
+
+// crawlFetcher adapts webfetch to crawl.Fetcher, which is the seam that keeps
+// net/http out of the crawler package.
+type crawlFetcher struct{ f *webfetch.Fetcher }
+
+func (a *crawlFetcher) Get(ctx context.Context, u string) (*crawl.Response, error) {
+	resp, err := a.f.Get(ctx, u)
+	if err != nil {
+		return nil, err
+	}
+	return &crawl.Response{URL: resp.URL, ContentType: resp.ContentType, Body: resp.Body}, nil
+}
+
+// factHashes stores each watch's last content hash as a config fact, so a
+// restart does not re-note an unchanged page. Same mechanism the feed reader
+// uses for its marks, and inspectable on /dash.
+type factHashes struct{ api ipc.CoreAPI }
+
+func hashKey(name string) string { return "crawl:hash:" + name }
+
+func (h *factHashes) LastHash(ctx context.Context, name string) (string, error) {
+	f, err := h.api.LatestFact(ctx, hashKey(name))
+	if err != nil {
+		// No hash yet is not an error: the watcher treats "" as "never read".
+		return "", nil
+	}
+	return f.Value, nil
+}
+
+func (h *factHashes) SetHash(ctx context.Context, name, hash string) error {
+	_, err := h.api.WriteFact(ctx, ipc.WriteFactReq{
+		Ts:         time.Now(),
+		Kind:       "config",
+		Key:        hashKey(name),
+		Value:      hash,
+		Source:     "poll:crawl",
+		Confidence: 1.0,
+	})
+	return err
+}
+
+// crawlEmbedder adapts router.Embedder for the watcher, embedding with
+// EmbedPassage (a page is text being searched FOR, and the e5 embedder is
+// asymmetric).
+func crawlEmbedder(emb router.Embedder) crawl.Embedder {
+	if emb == nil {
+		return nil
+	}
+	return passageEmbedder{emb}
+}

cmd/mavend/crawls_test.go

@@ -0,0 +1,186 @@
+package main
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+
+	"github.com/kami/maven/internal/config"
+	"github.com/kami/maven/internal/crawl"
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/phraser"
+	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/voice"
+)
+
+// The default config reads nothing. This is the whole "off unless configured"
+// contract for the crawler, asserted at the wiring level rather than trusted.
+func TestCrawlOffByDefault(t *testing.T) {
+	cfg := &config.Config{}
+	if c := newCrawler(cfg); c != nil {
+		t.Error("newCrawler with no crawl block returned a crawler")
+	}
+	if c := onDemandCrawler(cfg); c != nil {
+		t.Error("onDemandCrawler with no crawl block returned a crawler")
+	}
+	if w := newCrawlWorker(nil, nil, nil, cfg); w != nil {
+		t.Error("newCrawlWorker with no crawl block returned a worker")
+	}
+	// Watches configured but on_demand off ⇒ the answer path still reads
+	// nothing: a timer over a fixed list is not permission for arbitrary URLs.
+	withWatch := &config.Config{Crawl: &config.CrawlConfig{
+		Watches: []config.CrawlWatchConfig{{Name: "p", URL: "https://example.org/p"}},
+	}}
+	if c := onDemandCrawler(withWatch); c != nil {
+		t.Error("onDemandCrawler honoured a watch list as on-demand permission")
+	}
+	if c := newCrawler(withWatch); c == nil {
+		t.Error("newCrawler returned nil for a configured watch")
+	}
+}
+
+// The wired fetcher must refuse a private address, because the crawler on this
+// box sits one hop from the whole homelab. Same guard the webfetch tests cover;
+// this asserts the daemon actually wires it.
+func TestCrawlerRefusesPrivateAddress(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "text/html")
+		w.Write([]byte("<html><body>secret</body></html>"))
+	}))
+	defer srv.Close()
+
+	c := newCrawler(&config.Config{Crawl: &config.CrawlConfig{OnDemand: true}})
+	if c == nil {
+		t.Fatal("newCrawler returned nil for an on-demand config")
+	}
+	if _, err := c.Page(context.Background(), srv.URL); err == nil {
+		t.Fatalf("reading %s succeeded; a loopback address must be refused", srv.URL)
+	}
+}
+
+func TestFactHashesRoundTrip(t *testing.T) {
+	ctx := context.Background()
+	st := newTestStore(t)
+	h := &factHashes{api: ipc.NewStoreAPI(st)}
+
+	got, err := h.LastHash(ctx, "page")
+	if err != nil {
+		t.Fatalf("LastHash on a fresh store: %v", err)
+	}
+	if got != "" {
+		t.Errorf("LastHash = %q, want empty for a never-read page", got)
+	}
+	if err := h.SetHash(ctx, "page", "deadbeef"); err != nil {
+		t.Fatalf("SetHash: %v", err)
+	}
+	got, err = h.LastHash(ctx, "page")
+	if err != nil {
+		t.Fatalf("LastHash: %v", err)
+	}
+	if got != "deadbeef" {
+		t.Errorf("LastHash = %q, want deadbeef", got)
+	}
+	if key := hashKey("page"); key != "crawl:hash:page" {
+		t.Errorf("hashKey = %q", key)
+	}
+}
+
+// stubCrawlFetcher serves one fixed page to every URL, so queryWeb can be
+// exercised without a network or an allowlist.
+type stubCrawlFetcher struct{ body, ctype string }
+
+func (s *stubCrawlFetcher) Get(_ context.Context, u string) (*crawl.Response, error) {
+	ct := s.ctype
+	if ct == "" {
+		ct = "text/html"
+	}
+	if strings.HasSuffix(u, "/robots.txt") {
+		return &crawl.Response{URL: u, ContentType: "text/plain", Body: []byte("")}, nil
+	}
+	return &crawl.Response{URL: u, ContentType: ct, Body: []byte(s.body)}, nil
+}
+
+func buildWebHandler(c *crawl.Crawler) *reactiveHandler {
+	return &reactiveHandler{
+		replier: voice.NewStubReplier(),
+		phraser: phraser.NewStub(),
+		crawler: c,
+	}
+}
+
+func askWeb(h *reactiveHandler, q string) (string, bool) {
+	return h.queryWeb(context.Background(), &queryTurn{
+		dec: router.Decision{Intent: router.IntentQuery, Utterance: q},
+	})
+}
+
+func TestQueryWebPassesWithoutAURL(t *testing.T) {
+	h := buildWebHandler(crawl.New(&stubCrawlFetcher{body: "<html><body>x</body></html>"}, crawl.Config{}))
+	if reply, ok := askWeb(h, "почему небо синее?"); ok {
+		t.Errorf("the web source claimed a question with no URL: %q", reply)
+	}
+}
+
+// Not configured is said out loud rather than falling through, so a small model
+// never invents a page's contents from its URL.
+func TestQueryWebSaysWhenNotConfigured(t *testing.T) {
+	h := buildWebHandler(nil)
+	reply, ok := askWeb(h, "посмотри https://example.org/page")
+	if !ok {
+		t.Fatal("the web source did not claim a question with a URL")
+	}
+	if !strings.Contains(reply, "не настроено") {
+		t.Errorf("reply = %q, want the not-configured answer", reply)
+	}
+}
+
+func TestQueryWebReadsThePage(t *testing.T) {
+	h := buildWebHandler(crawl.New(&stubCrawlFetcher{
+		body: "<html><head><title>Заголовок</title></head><body><p>текст страницы</p></body></html>",
+	}, crawl.Config{}))
+	reply, ok := askWeb(h, "посмотри https://example.org/page — что там?")
+	if !ok {
+		t.Fatal("the web source did not claim a question with a URL")
+	}
+	if !strings.Contains(reply, "текст страницы") {
+		t.Errorf("reply = %q, want the page text read back", reply)
+	}
+}
+
+func TestQueryWebRefusesNonHTML(t *testing.T) {
+	h := buildWebHandler(crawl.New(&stubCrawlFetcher{
+		body: "\x00\x01binary", ctype: "application/octet-stream",
+	}, crawl.Config{}))
+	reply, ok := askWeb(h, "почитай https://example.org/blob.bin")
+	if !ok {
+		t.Fatal("the web source did not claim a question with a URL")
+	}
+	if !strings.Contains(reply, "не получилось") {
+		t.Errorf("reply = %q, want the read-failed answer", reply)
+	}
+}
+
+// robots.txt is honoured on the answer path too, and she says so instead of
+// reporting a generic failure.
+func TestQueryWebObeysRobots(t *testing.T) {
+	h := buildWebHandler(crawl.New(&robotsDenyFetcher{}, crawl.Config{}))
+	reply, ok := askWeb(h, "посмотри https://example.org/private")
+	if !ok {
+		t.Fatal("the web source did not claim a question with a URL")
+	}
+	if !strings.Contains(reply, "robots.txt") {
+		t.Errorf("reply = %q, want the robots answer", reply)
+	}
+}
+
+type robotsDenyFetcher struct{}
+
+func (robotsDenyFetcher) Get(_ context.Context, u string) (*crawl.Response, error) {
+	if strings.HasSuffix(u, "/robots.txt") {
+		return &crawl.Response{URL: u, ContentType: "text/plain",
+			Body: []byte("User-agent: *\nDisallow: /private\n")}, nil
+	}
+	return &crawl.Response{URL: u, ContentType: "text/html", Body: []byte("<html>nope</html>")}, nil
+}

cmd/mavend/feeds.go

@@ -0,0 +1,182 @@
+// mavend/feeds.go — the driver for RSS/Atom reading (Vikunja #258,
+// docs/plans/13-rss-news-feeds.md). The reader itself is pure and lives in
+// internal/rss; this is the impure half: a ticker, the guarded fetcher, and the
+// two adapters that let a pure package talk to the store.
+//
+// Why in-core rather than its own daemon like mavmaild and mavpoll: those two
+// hold a CREDENTIAL (an IMAP password, a zenmoney token), and the reason they
+// are separate processes is that core must never see it. A feed URL is public,
+// there is no secret to isolate, and a whole extra binary and compose service
+// would buy nothing. The other half of the mavpoll precedent — off unless
+// configured — is kept: no `feeds` block, no poller, no outbound request.
+//
+// It is its own goroutine, not a step on the tick: the tick has a delivery
+// deadline behind it, and a feed read is a network round-trip that nobody is
+// waiting on.
+//
+// Nothing here dispatches. A feed that announced itself would be a nag, so the
+// only output is notes with source "rss:<feed>", which the answer path reads
+// when he asks ("что нового в лентах?" — see queryFeeds in actions_query.go).
+package main
+
+import (
+	"context"
+	"log"
+	"net/url"
+	"time"
+
+	"github.com/kami/maven/internal/config"
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/rss"
+	"github.com/kami/maven/internal/webfetch"
+)
+
+// feedWorker — ticker + poller.
+type feedWorker struct {
+	poller   *rss.Poller
+	interval time.Duration
+}
+
+// feedTickInterval — how often the worker asks the poller what is due. Per-feed
+// cadence is the poller's business; this is just the granularity.
+const feedTickInterval = 5 * time.Minute
+
+// newFeedWorker wires feed reading, or returns nil when it must not run:
+// no `feeds` block (the normal case), or nothing valid in it. Every caller
+// checks for nil.
+func newFeedWorker(api ipc.CoreAPI, emb router.Embedder, cfg *config.Config) *feedWorker {
+	if cfg.Feeds == nil {
+		return nil
+	}
+	fc := cfg.Feeds
+
+	feeds := make([]rss.FeedConfig, 0, len(fc.Sources))
+	hosts := append([]string(nil), fc.AllowHosts...)
+	for _, s := range fc.Sources {
+		feeds = append(feeds, rss.FeedConfig{
+			Name:     s.Name,
+			URL:      s.URL,
+			Category: s.Category,
+			Interval: time.Duration(s.Interval),
+			Include:  s.Include,
+			Exclude:  s.Exclude,
+		})
+		// Each configured feed's own host is allowed. The allowlist is then
+		// exactly "the feeds he asked for", so a redirect off to somewhere else
+		// is refused by the fetcher rather than followed.
+		if u, err := url.Parse(s.URL); err == nil && u.Hostname() != "" {
+			hosts = append(hosts, u.Hostname())
+		}
+	}
+
+	fetcher := webfetch.New(webfetch.Config{
+		AllowHosts: hosts,
+		Timeout:    time.Duration(fc.Timeout),
+		MaxBytes:   fc.MaxBytes,
+	})
+	poller := rss.NewPoller(feeds, &feedFetcher{f: fetcher}, api, &factMarks{api: api},
+		embedderFor(emb), nil, rss.Config{
+			DefaultInterval: time.Duration(fc.PollInterval),
+			MaxItems:        fc.MaxItems,
+			MaxAge:          time.Duration(fc.MaxAge),
+		})
+	if poller == nil {
+		log.Printf("feeds: configured but nothing pollable — feed reading disabled")
+		return nil
+	}
+	log.Printf("feeds: reading %d feed(s), checking what is due every %s", len(feeds), feedTickInterval)
+	return &feedWorker{poller: poller, interval: feedTickInterval}
+}
+
+// run polls what is due until ctx is canceled. The first round runs immediately
+// so a restart does not blind her for the first interval; it writes notes only,
+// so an early round cannot startle anyone.
+func (w *feedWorker) run(ctx context.Context) {
+	w.poller.PollDue(ctx, time.Now())
+	t := time.NewTicker(w.interval)
+	defer t.Stop()
+	for {
+		select {
+		case <-ctx.Done():
+			return
+		case now := <-t.C:
+			w.poller.PollDue(ctx, now)
+		}
+	}
+}
+
+// embedderOf — the voice wiring's embedder, or nil when voice is not wired.
+// Feed notes are embedded with the SAME model the rest of the store uses, or not
+// at all; a second embedder would write vectors nothing can search.
+func embedderOf(w *voiceWiring) router.Embedder {
+	if w == nil {
+		return nil
+	}
+	return w.embedder
+}
+
+// feedFetcher adapts webfetch to rss.Fetcher — the pure package names the two
+// fields it needs and stays free of net/http.
+type feedFetcher struct{ f *webfetch.Fetcher }
+
+func (a *feedFetcher) Get(ctx context.Context, u string) (*rss.Body, error) {
+	resp, err := a.f.Get(ctx, u)
+	if err != nil {
+		return nil, err
+	}
+	return &rss.Body{Bytes: resp.Body}, nil
+}
+
+// factMarks stores "how far this feed was read" as a config fact, the same
+// mechanism the plan named and the same one the pattern tick uses for its own
+// bookkeeping. Durable, inspectable on /dash, and cheap.
+type factMarks struct{ api ipc.CoreAPI }
+
+func markKey(feed string) string { return "rss:latest:" + feed }
+
+func (m *factMarks) LastMark(ctx context.Context, feed string) (time.Time, error) {
+	f, err := m.api.LatestFact(ctx, markKey(feed))
+	if err != nil {
+		// No mark yet is not an error worth propagating: the poller treats a
+		// zero time as a cold start.
+		return time.Time{}, nil
+	}
+	t, err := time.Parse(time.RFC3339, f.Value)
+	if err != nil {
+		return time.Time{}, nil
+	}
+	return t, nil
+}
+
+func (m *factMarks) SetMark(ctx context.Context, feed string, at time.Time) error {
+	_, err := m.api.WriteFact(ctx, ipc.WriteFactReq{
+		Ts:         time.Now(),
+		Kind:       "config",
+		Key:        markKey(feed),
+		Value:      at.UTC().Format(time.RFC3339),
+		Source:     "poll:rss",
+		Confidence: 1.0,
+	})
+	return err
+}
+
+// embedderFor adapts router.Embedder to rss.Embedder, and returns nil when
+// there is none — a note without a vector is still a note the recent-notes path
+// can read.
+//
+// EmbedPassage, not Embed: a feed item is text being searched FOR, and the e5
+// embedder is asymmetric. Getting this backwards makes the item unfindable by
+// the question that should have matched it.
+func embedderFor(emb router.Embedder) rss.Embedder {
+	if emb == nil {
+		return nil
+	}
+	return passageEmbedder{emb}
+}
+
+type passageEmbedder struct{ e router.Embedder }
+
+func (p passageEmbedder) Embed(ctx context.Context, text string) ([]float32, error) {
+	return router.EmbedPassage(ctx, p.e, text)
+}

cmd/mavend/feeds_test.go

@@ -0,0 +1,177 @@
+package main
+
+import (
+	"context"
+	"strings"
+	"testing"
+	"time"
+
+	"github.com/kami/maven/internal/config"
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/phraser"
+	"github.com/kami/maven/internal/router"
+	"github.com/kami/maven/internal/rss"
+	"github.com/kami/maven/internal/voice"
+)
+
+// buildFeedHandler — a handler with the given feed notes already stored. No
+// embedder: the feed source answers from recent notes by source, which is what
+// makes it work for notes written before an embedder existed.
+func buildFeedHandler(t *testing.T, feedsOn bool, notes ...ipc.Note) *reactiveHandler {
+	t.Helper()
+	ctx := context.Background()
+	st := newTestStore(t)
+	now := time.Now()
+	for i, n := range notes {
+		ts := now.Add(time.Duration(i) * time.Minute)
+		if _, err := st.WriteNote(ctx, ts, n.Text, nil, n.Source); err != nil {
+			t.Fatalf("WriteNote: %v", err)
+		}
+	}
+	return &reactiveHandler{
+		api:      ipc.NewStoreAPI(st),
+		replier:  voice.NewStubReplier(),
+		phraser:  phraser.NewStub(),
+		now:      func() time.Time { return now },
+		feedsOn:  feedsOn,
+		embedder: nil,
+	}
+}
+
+func askFeeds(t *testing.T, h *reactiveHandler, q string) (string, bool) {
+	t.Helper()
+	return h.queryFeeds(context.Background(), &queryTurn{
+		dec: router.Decision{Intent: router.IntentQuery, Utterance: q},
+	})
+}
+
+func TestQueryFeedsReadsFeedNotes(t *testing.T) {
+	h := buildFeedHandler(t, true,
+		ipc.Note{Text: "Новая уязвимость в ядре [технологии]\nпатч вышел\nhttps://example.org/a", Source: "rss:habr"},
+		ipc.Note{Text: "что-то он сам сказал", Source: "tap:voice"},
+	)
+	reply, ok := askFeeds(t, h, "что нового в лентах?")
+	if !ok {
+		t.Fatal("the feed source did not claim the question")
+	}
+	if !strings.Contains(reply, "уязвимость") {
+		t.Errorf("reply = %q, want the headline", reply)
+	}
+	if strings.Contains(reply, "он сам сказал") {
+		t.Errorf("a note he dictated leaked into the feed answer: %q", reply)
+	}
+	// She reads the headline, not the summary and not the URL.
+	if strings.Contains(reply, "https://") || strings.Contains(reply, "патч вышел") {
+		t.Errorf("reply = %q, want the title line only", reply)
+	}
+}
+
+func TestQueryFeedsByCategory(t *testing.T) {
+	h := buildFeedHandler(t, true,
+		ipc.Note{Text: "Релиз ядра [технологии]", Source: "rss:habr"},
+		ipc.Note{Text: "Выборы отложены [политика]", Source: "rss:news"},
+	)
+	reply, ok := askFeeds(t, h, "что нового по технологиям?")
+	if !ok {
+		t.Fatal("not claimed")
+	}
+	if !strings.Contains(reply, "ядра") || strings.Contains(reply, "Выборы") {
+		t.Fatalf("reply = %q, want only the технологии item", reply)
+	}
+	reply, _ = askFeeds(t, h, "что нового по спорту?")
+	if !strings.Contains(reply, "ничего") {
+		t.Fatalf("reply = %q, want an honest empty answer for an unread category", reply)
+	}
+}
+
+// "не настроены" and "ничего нового" are different truths, and neither may be
+// answered by the model inventing a bulletin.
+func TestQueryFeedsOffAndEmptyDiffer(t *testing.T) {
+	off := buildFeedHandler(t, false)
+	reply, ok := askFeeds(t, off, "что нового?")
+	if !ok || !strings.Contains(reply, "не настроены") {
+		t.Fatalf("feeds off: reply = %q, ok = %v", reply, ok)
+	}
+	on := buildFeedHandler(t, true)
+	reply, ok = askFeeds(t, on, "что нового?")
+	if !ok || !strings.Contains(reply, "ничего нового") {
+		t.Fatalf("feeds on but empty: reply = %q, ok = %v", reply, ok)
+	}
+}
+
+func TestQueryFeedsPassesOnANonFeedQuestion(t *testing.T) {
+	h := buildFeedHandler(t, true)
+	if reply, ok := askFeeds(t, h, "напомни полить цветы"); ok {
+		t.Fatalf("claimed an unrelated question with %q", reply)
+	}
+}
+
+// The mark is what stops a restart from re-noting yesterday's headlines, so the
+// fact round-trip is worth a test of its own.
+func TestFactMarksRoundTrip(t *testing.T) {
+	st := newTestStore(t)
+	m := &factMarks{api: ipc.NewStoreAPI(st)}
+	ctx := context.Background()
+
+	at, err := m.LastMark(ctx, "habr")
+	if err != nil || !at.IsZero() {
+		t.Fatalf("no mark yet: got %v, %v — want zero time and no error", at, err)
+	}
+	want := time.Date(2026, 7, 28, 10, 0, 0, 0, time.UTC)
+	if err := m.SetMark(ctx, "habr", want); err != nil {
+		t.Fatal(err)
+	}
+	got, err := m.LastMark(ctx, "habr")
+	if err != nil {
+		t.Fatal(err)
+	}
+	if !got.Equal(want) {
+		t.Fatalf("mark = %v, want %v", got, want)
+	}
+}
+
+// Off unless configured, checked at the wiring seam: no `feeds` block ⇒ no
+// worker ⇒ no outbound request is possible.
+func TestNewFeedWorkerOffByDefault(t *testing.T) {
+	st := newTestStore(t)
+	api := ipc.NewStoreAPI(st)
+	if w := newFeedWorker(api, nil, &config.Config{}); w != nil {
+		t.Fatal("a config with no feeds block wired a feed worker")
+	}
+	// An empty sources list is normalised to "off" by config.Load; the worker
+	// refuses it too, so a hand-built Config cannot switch it on by accident.
+	if w := newFeedWorker(api, nil, &config.Config{Feeds: &config.FeedsConfig{}}); w != nil {
+		t.Fatal("an empty sources list wired a feed worker")
+	}
+	cfg := &config.Config{Feeds: &config.FeedsConfig{Sources: []config.FeedSourceConfig{
+		{Name: "habr", URL: "https://example.org/rss"},
+	}}}
+	w := newFeedWorker(api, nil, cfg)
+	if w == nil {
+		t.Fatal("a configured feed did not wire a worker")
+	}
+	if got := w.poller.Feeds(); len(got) != 1 || got[0].Name != "habr" {
+		t.Fatalf("feeds = %+v", got)
+	}
+}
+
+// The fetcher the worker builds must be allowlisted to the configured feeds and
+// nothing else — the crawler's SSRF guards are only worth as much as the
+// allowlist handed to them.
+func TestFeedWorkerFetcherIsAllowlisted(t *testing.T) {
+	cfg := &config.Config{Feeds: &config.FeedsConfig{Sources: []config.FeedSourceConfig{
+		{Name: "habr", URL: "https://feeds.example.org/rss"},
+	}}}
+	w := newFeedWorker(ipc.NewStoreAPI(newTestStore(t)), nil, cfg)
+	if w == nil {
+		t.Fatal("no worker")
+	}
+	// PollFeed goes through the guarded fetcher; a feed URL pointing at the box
+	// itself must fail rather than be read.
+	_, err := w.poller.PollFeed(context.Background(), rss.FeedConfig{
+		Name: "evil", URL: "http://127.0.0.1:9100/mcp",
+	}, time.Now())
+	if err == nil {
+		t.Fatal("the poller fetched a private address")
+	}
+}

cmd/mavend/mail.go

@@ -0,0 +1,158 @@
+// mavend/mail.go — core's half of the email reader (Vikunja #246,
+// docs/plans/01-email-reader.md).
+//
+// The split: cmd/mavmaild holds the IMAP credential, connects to the mailbox
+// and converts messages to plaintext; it hands each message to core over
+// ipc.MethodIngestMail. Core runs the extraction on the resident model —
+// llama-server lives in this process, spawned by the phraser — and writes what
+// comes back through the one task intake seam.
+//
+// What this file may produce is exactly one thing: rows in `tasks` with status
+// "candidate". No fact, no reminder, no note, no nudge, no calendar event. A
+// 1.7B misreading a mail can therefore put a wrong line on a review page and
+// nothing else; it can never make Maven speak, and it can never make her
+// recite something out of an advert as true.
+//
+// Off unless configured twice over: no `email` block in mavend.json ⇒ the IPC
+// method does not exist; no llama-server phraser ⇒ same. A reader pointed at a
+// core that is not set up for mail gets ErrUnknownMethod rather than silence.
+package main
+
+import (
+	"context"
+	"fmt"
+	"log"
+	"time"
+
+	"github.com/kami/maven/internal/config"
+	"github.com/kami/maven/internal/email"
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/llm"
+	"github.com/kami/maven/internal/phraser"
+	"github.com/kami/maven/internal/store"
+)
+
+// evidenceMaxChars — how much of the subject line is kept as a candidate's
+// evidence. Enough to recognise the mail on /tasks, not enough to turn the task
+// list into a copy of his mailbox.
+const evidenceMaxChars = 160
+
+// mailIntake — extraction + capture for one message at a time.
+type mailIntake struct {
+	st      *store.Store
+	ex      *email.Extractor
+	timeout time.Duration
+	now     func() time.Time
+}
+
+// newMailIntake returns nil when mail ingestion must not be available, which is
+// the default. Both preconditions are real:
+//
+//   - no cfg.Email ⇒ not configured, and a capability is off unless configured;
+//   - no llama-server phraser ⇒ nothing to extract with. There is deliberately
+//     no keyword fallback: "the subject line became a task" is not extraction,
+//     it is a mailbox rendered as a to-do list, and it would fill the review
+//     page faster than he could clear it.
+func newMailIntake(st *store.Store, phr phraser.Phraser, cfg *config.Config) *mailIntake {
+	if cfg.Email == nil {
+		return nil
+	}
+	lp, ok := phr.(*phraser.LLMPhraser)
+	if !ok {
+		log.Printf("mail intake: configured but no llama-server phraser — mail ingestion disabled")
+		return nil
+	}
+	timeout := time.Duration(cfg.Email.Timeout)
+	if timeout <= 0 {
+		timeout = config.DefaultEmailTimeout
+	}
+	ex := email.NewExtractor(llm.New(lp.BaseURL(), timeout), cfg.Email.MaxTasks, contextBlockFn(cfg, time.Now))
+	log.Printf("mail intake: enabled (max %d candidates per message, timeout %s)", cfg.Email.MaxTasks, timeout)
+	return &mailIntake{st: st, ex: ex, timeout: timeout, now: time.Now}
+}
+
+// ingest handles one ipc.MethodIngestMail call.
+//
+// Junk and empty messages are answered Skipped without touching the model — the
+// reader's header filter is what keeps the resident model off newsletters.
+//
+// Every candidate is captured with Status "candidate", Source "email:<mailbox>"
+// and the subject as Evidence. CaptureTask dedupes on normalised text among
+// live rows, so a mailbox re-read after a restart produces Created=0 rather
+// than a second copy of every task.
+func (m *mailIntake) ingest(ctx context.Context, req ipc.IngestMailReq) (ipc.IngestMailResp, error) {
+	msg := email.Message{
+		UID:     req.UID,
+		From:    req.From,
+		Subject: req.Subject,
+		Date:    req.Date,
+		Body:    req.Body,
+		Junk:    req.Junk,
+	}
+	if msg.Junk || (msg.Subject == "" && msg.Body == "") {
+		return ipc.IngestMailResp{Skipped: true}, nil
+	}
+
+	ctx, cancel := context.WithTimeout(ctx, m.timeout)
+	defer cancel()
+	cands, err := m.ex.Extract(ctx, msg)
+	if err != nil {
+		// The error from internal/email never carries mail text; keep it that way
+		// by not adding the subject here.
+		return ipc.IngestMailResp{}, fmt.Errorf("mail intake: uid %d: %w", req.UID, err)
+	}
+	if len(cands) == 0 {
+		return ipc.IngestMailResp{}, nil
+	}
+
+	source := email.SourcePrefix + req.Mailbox
+	evidence := truncateRunes(req.Subject, evidenceMaxChars)
+	now := m.now()
+	var resp ipc.IngestMailResp
+	for _, c := range cands {
+		t := store.Task{
+			CreatedTs: now,
+			Text:      c.Text,
+			Source:    source,
+			Evidence:  evidence,
+			// The one status this path may ever write. Anything Maven derived from
+			// something she read is a suggestion until he confirms it on /tasks.
+			Status: store.TaskCandidate,
+		}
+		if due, ok := email.ParseDue(c.Due); ok {
+			t.Due = &due
+		}
+		id, created, err := m.st.CaptureTask(ctx, t)
+		if err != nil {
+			return resp, fmt.Errorf("mail intake: capture: %w", err)
+		}
+		resp.TaskIDs = append(resp.TaskIDs, id)
+		if created {
+			resp.Created++
+		}
+	}
+	// Counts only: the log line names the mailbox and the UID, never the subject,
+	// the sender or the task text. Reviewing a candidate is what /tasks is for.
+	log.Printf("mail intake: %s uid %d → %d candidate(s), %d new", source, req.UID, len(cands), resp.Created)
+	return resp, nil
+}
+
+// wireMailIntake installs the IPC hook, or leaves it nil so the method reports
+// ErrUnknownMethod. Called on both startup paths (unlocked boot and passkey
+// unlock) so mail behaves the same either way.
+func wireMailIntake(srv *ipc.Server, st *store.Store, phr phraser.Phraser, cfg *config.Config) {
+	mi := newMailIntake(st, phr, cfg)
+	if mi == nil {
+		return
+	}
+	srv.IngestMailFn = mi.ingest
+}
+
+// truncateRunes cuts a string to n runes, marking the cut.
+func truncateRunes(s string, n int) string {
+	r := []rune(s)
+	if len(r) <= n {
+		return s
+	}
+	return string(r[:n]) + "…"
+}

cmd/mavend/mail_test.go

@@ -0,0 +1,182 @@
+package main
+
+import (
+	"context"
+	"strings"
+	"testing"
+	"time"
+
+	"github.com/kami/maven/internal/config"
+	"github.com/kami/maven/internal/email"
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/llm"
+	"github.com/kami/maven/internal/store"
+)
+
+// mailLLM — a canned extraction reply.
+type mailLLM struct {
+	reply string
+	calls int
+}
+
+func (m *mailLLM) Complete(_ context.Context, _ llm.Req) (string, error) {
+	m.calls++
+	return m.reply, nil
+}
+
+func newTestIntake(t *testing.T, reply string) (*mailIntake, *store.Store, *mailLLM) {
+	t.Helper()
+	st := newTestStore(t)
+	fake := &mailLLM{reply: reply}
+	return &mailIntake{
+		st:      st,
+		ex:      email.NewExtractor(fake, 0, nil),
+		timeout: 5 * time.Second,
+		now:     func() time.Time { return time.Date(2026, 8, 1, 10, 0, 0, 0, time.UTC) },
+	}, st, fake
+}
+
+func ingestReq() ipc.IngestMailReq {
+	return ipc.IngestMailReq{
+		Mailbox: "INBOX", UID: 42,
+		From:    "billing@isp.example",
+		Subject: "Счёт за интернет",
+		Body:    "Оплатите счёт до 5 августа.",
+	}
+}
+
+// The one property that matters: a mail-derived task is a candidate, attributed
+// to the mailbox, with the subject as reviewable evidence — and nothing else is
+// written.
+func TestIngestCapturesCandidates(t *testing.T) {
+	mi, st, _ := newTestIntake(t, `[{"text":"оплатить счёт за интернет","due":"2026-08-05"}]`)
+	resp, err := mi.ingest(context.Background(), ingestReq())
+	if err != nil {
+		t.Fatalf("ingest: %v", err)
+	}
+	if resp.Created != 1 || len(resp.TaskIDs) != 1 {
+		t.Fatalf("resp = %+v, want one created task", resp)
+	}
+	tasks, err := st.ListTasks(context.Background(), "")
+	if err != nil {
+		t.Fatalf("list: %v", err)
+	}
+	if len(tasks) != 1 {
+		t.Fatalf("got %d tasks, want 1", len(tasks))
+	}
+	got := tasks[0]
+	if got.Status != store.TaskCandidate {
+		t.Errorf("status = %q, want %q — mail may only produce candidates", got.Status, store.TaskCandidate)
+	}
+	if got.Source != "email:INBOX" {
+		t.Errorf("source = %q, want email:INBOX", got.Source)
+	}
+	if got.Evidence != "Счёт за интернет" {
+		t.Errorf("evidence = %q, want the subject line", got.Evidence)
+	}
+	if got.Due == nil || got.Due.Format("2006-01-02") != "2026-08-05" {
+		t.Errorf("due = %v, want 2026-08-05", got.Due)
+	}
+	// Nothing else may have been written: no reminder, no fact.
+	rem, err := st.ListReminders(context.Background(), 10)
+	if err != nil {
+		t.Fatalf("list reminders: %v", err)
+	}
+	if len(rem) != 0 {
+		t.Errorf("mail created %d reminders; a misread mail must never be able to fire", len(rem))
+	}
+}
+
+// Re-reading a mailbox must not grow the list — CaptureTask dedupes among live
+// rows, and the intake relies on exactly that.
+func TestIngestSameMailTwiceIsIdempotent(t *testing.T) {
+	mi, st, _ := newTestIntake(t, `[{"text":"оплатить счёт","due":""}]`)
+	if _, err := mi.ingest(context.Background(), ingestReq()); err != nil {
+		t.Fatalf("first ingest: %v", err)
+	}
+	resp, err := mi.ingest(context.Background(), ingestReq())
+	if err != nil {
+		t.Fatalf("second ingest: %v", err)
+	}
+	if resp.Created != 0 || len(resp.TaskIDs) != 1 {
+		t.Errorf("resp = %+v, want the existing row and Created=0", resp)
+	}
+	tasks, _ := st.ListTasks(context.Background(), "")
+	if len(tasks) != 1 {
+		t.Errorf("got %d tasks after two reads, want 1", len(tasks))
+	}
+}
+
+func TestIngestJunkSkipsTheModel(t *testing.T) {
+	mi, st, fake := newTestIntake(t, `[{"text":"купить со скидкой","due":""}]`)
+	req := ingestReq()
+	req.Junk = true
+	resp, err := mi.ingest(context.Background(), req)
+	if err != nil {
+		t.Fatalf("ingest: %v", err)
+	}
+	if !resp.Skipped || resp.Created != 0 {
+		t.Errorf("resp = %+v, want skipped", resp)
+	}
+	if fake.calls != 0 {
+		t.Errorf("model called %d times for junk, want 0", fake.calls)
+	}
+	if tasks, _ := st.ListTasks(context.Background(), ""); len(tasks) != 0 {
+		t.Errorf("junk produced %d tasks, want 0", len(tasks))
+	}
+}
+
+func TestIngestEmptyMessageSkipped(t *testing.T) {
+	mi, _, fake := newTestIntake(t, "[]")
+	resp, err := mi.ingest(context.Background(), ipc.IngestMailReq{Mailbox: "INBOX", UID: 1})
+	if err != nil || !resp.Skipped {
+		t.Fatalf("resp = %+v, err = %v; want skipped", resp, err)
+	}
+	if fake.calls != 0 {
+		t.Errorf("model called %d times for an empty message, want 0", fake.calls)
+	}
+}
+
+func TestIngestNoTasksWritesNothing(t *testing.T) {
+	mi, st, _ := newTestIntake(t, "[]")
+	resp, err := mi.ingest(context.Background(), ingestReq())
+	if err != nil {
+		t.Fatalf("ingest: %v", err)
+	}
+	if resp.Created != 0 || len(resp.TaskIDs) != 0 || resp.Skipped {
+		t.Errorf("resp = %+v, want nothing captured and not skipped", resp)
+	}
+	if tasks, _ := st.ListTasks(context.Background(), ""); len(tasks) != 0 {
+		t.Errorf("got %d tasks, want 0", len(tasks))
+	}
+}
+
+func TestIngestTruncatesEvidence(t *testing.T) {
+	mi, st, _ := newTestIntake(t, `[{"text":"дело","due":""}]`)
+	req := ingestReq()
+	req.Subject = strings.Repeat("щ", 400)
+	if _, err := mi.ingest(context.Background(), req); err != nil {
+		t.Fatalf("ingest: %v", err)
+	}
+	tasks, _ := st.ListTasks(context.Background(), "")
+	if len(tasks) != 1 {
+		t.Fatalf("got %d tasks, want 1", len(tasks))
+	}
+	if n := len([]rune(tasks[0].Evidence)); n > evidenceMaxChars+1 {
+		t.Errorf("evidence kept %d runes, want ≤ %d", n, evidenceMaxChars)
+	}
+}
+
+// Off unless configured: no email block ⇒ no intake, so the IPC method does not
+// exist at all.
+func TestNewMailIntakeOffWithoutConfig(t *testing.T) {
+	st := newTestStore(t)
+	if mi := newMailIntake(st, nil, &config.Config{}); mi != nil {
+		t.Error("no email block must mean no mail intake")
+	}
+	// Configured but with a non-LLM phraser: still off — there is no fallback
+	// extraction, by design.
+	if mi := newMailIntake(st, nil, &config.Config{Email: &config.EmailConfig{}}); mi != nil {
+		t.Error("without a llama-server phraser there is nothing to extract with")
+	}
+}

cmd/mavend/main.go

@@ -170,6 +170,8 @@ func run(args []string) error {
 		eco        *ecosystemWiring
 		factWorker *factEnrichmentWorker
 		evalWorker *memoryEvalWorker // nil ⇒ memory evaluation off (the default)
+		feedWkr    *feedWorker       // nil ⇒ no feed is read (the default)
+		crawlWkr   *crawlWorker      // nil ⇒ no page is watched (the default)
 	)
 
 	if !locked {
@@ -265,6 +267,8 @@ func run(args []string) error {
 		tl = newTickLoop(st, gatherer, dispatcher, phr, rules, tickInterval, repeatInterval, autotuneInterval, cfg.Digest, routinesFromConfig(cfg.Routines), config.MorningRoutinesFromConfig(cfg.MorningRoutines), cfg.PatternProposals)
 		factWorker = newFactEnrichmentWorker(st, eco, time.Duration(cfg.FactEnrichmentInterval))
 		evalWorker = newMemoryEvalWorker(st, phr, cfg)
+		feedWkr = newFeedWorker(ipc.NewStoreAPI(st), embedderOf(voiceW), cfg)
+		crawlWkr = newCrawlWorker(newCrawler(cfg), ipc.NewStoreAPI(st), embedderOf(voiceW), cfg)
 
 		coreAPI = &daemonAPI{
 			CoreAPI:          ipc.NewStoreAPI(st),
@@ -320,6 +324,13 @@ func run(args []string) error {
 
 	srv.StepUp = func(ctx context.Context) error { return passkeySess.Assert(ctx, auth.Scope{}) }
 
+	// Mail ingestion (Vikunja #246): the hook stays nil unless an email block is
+	// configured and there is a llama-server to extract with, in which case
+	// ipc.MethodIngestMail reports ErrUnknownMethod.
+	if !locked {
+		wireMailIntake(srv, st, phr, cfg)
+	}
+
 	// WrapKeyFn — wraps the env key with a passkey credential public key and
 	// persists the wrapped blob. Only wired when the daemon has the key in
 	// memory (env key mode). Called by mavweb after passkey enrollment.
@@ -444,6 +455,8 @@ func run(args []string) error {
 			tl = newTickLoop(st, gatherer, dispatcher, phr, rules, tickInterval, repeatInterval, autotuneInterval, cfg.Digest, routinesFromConfig(cfg.Routines), config.MorningRoutinesFromConfig(cfg.MorningRoutines), cfg.PatternProposals)
 			factWorker = newFactEnrichmentWorker(st, eco, time.Duration(cfg.FactEnrichmentInterval))
 			evalWorker = newMemoryEvalWorker(st, phr, cfg)
+			feedWkr = newFeedWorker(ipc.NewStoreAPI(st), embedderOf(voiceW), cfg)
+			crawlWkr = newCrawlWorker(newCrawler(cfg), ipc.NewStoreAPI(st), embedderOf(voiceW), cfg)
 
 			// Swap the CoreAPI from the locked placeholder to the real store adapter.
 			newAPI := &daemonAPI{
@@ -457,6 +470,7 @@ func run(args []string) error {
 			}
 			srv.SetAPI(newAPI)
 			srv.Check = (&auth.Gate{Enrollment: auth.NewFloorEnrollment(), Session: passkeySess}).Check
+			wireMailIntake(srv, st, phr, cfg)
 
 			// Start voice server.
 			if voiceW != nil {
@@ -488,6 +502,20 @@ func run(args []string) error {
 				}()
 			}
 
+			// Start feed reading (nil unless configured).
+			if feedWkr != nil {
+				go func() {
+					feedWkr.run(ctx)
+				}()
+			}
+
+			// Start the watched-page crawls (nil unless configured).
+			if crawlWkr != nil {
+				go func() {
+					crawlWkr.run(ctx)
+				}()
+			}
+
 			dl.unlock()
 			log.Printf("mavend: unlocked via passkey assertion")
 			return nil
@@ -533,6 +561,20 @@ func run(args []string) error {
 				evalWorker.run(ctx)
 			}()
 		}
+		if feedWkr != nil {
+			wg.Add(1)
+			go func() {
+				defer wg.Done()
+				feedWkr.run(ctx)
+			}()
+		}
+		if crawlWkr != nil {
+			wg.Add(1)
+			go func() {
+				defer wg.Done()
+				crawlWkr.run(ctx)
+			}()
+		}
 	}
 
 	<-ctx.Done()

cmd/mavend/voice.go

@@ -52,6 +52,7 @@ import (
 	"time"
 
 	"github.com/kami/maven/internal/audio"
+	"github.com/kami/maven/internal/crawl"
 	"github.com/kami/maven/internal/dialogue"
 	"github.com/kami/maven/internal/ipc"
 	"github.com/kami/maven/internal/memory"
@@ -82,6 +83,15 @@ type reactiveHandler struct {
 	replier  voice.Replier
 	now      func() time.Time
 
+	// crawler reads a web page he names out loud (queryWeb). nil ⇒ on-demand
+	// page reading is off, which is the default: no `crawl` block, no fetch.
+	crawler *crawl.Crawler
+
+	// feedsOn — whether any RSS feed is configured (config.Feeds). It changes
+	// only what she SAYS when asked and nothing is there: "ленты не настроены"
+	// instead of "ничего нового", which are different truths.
+	feedsOn bool
+
 	weatherProvider weather.Provider
 	weatherLocation string // default location for weather queries
 

cmd/mavend/voicewire.go

@@ -201,16 +201,20 @@ func wireVoice(cfg *config.Config, coreAPI ipc.CoreAPI, phr phraser.Phraser, mem
 
 	// ----- the handler (the reactive path; closes over stt / tts / router / coreAPI / memory) -----
 	h := &reactiveHandler{
-		stt:              transcriber,
-		tts:              synthesizer,
-		router:           rtr,
-		embedder:         emb,
-		api:              coreAPI,
-		tools:            exec,
-		matcher:          matcher,
-		replier:          replier,
-		phraser:          phr,
-		now:              time.Now,
+		stt:      transcriber,
+		tts:      synthesizer,
+		router:   rtr,
+		embedder: emb,
+		api:      coreAPI,
+		tools:    exec,
+		matcher:  matcher,
+		replier:  replier,
+		phraser:  phr,
+		now:      time.Now,
+		feedsOn:  cfg.Feeds != nil,
+		// nil unless `crawl.on_demand` is on: reading a page he names is a
+		// capability, and capabilities are off unless configured.
+		crawler:          onDemandCrawler(cfg),
 		weatherProvider:  weatherProvider,
 		weatherLocation:  weatherLocation,
 		memStore:         memStore,

cmd/mavmaild/main.go

@@ -0,0 +1,332 @@
+// mavmaild — the mail reader module (Vikunja #246,
+// docs/plans/01-email-reader.md).
+//
+// Every so often it opens one IMAP mailbox read-only, fetches the messages it
+// has not read yet, and hands each one to core over ipc.MethodIngestMail. Core
+// runs the extraction on the resident model and writes what comes back as task
+// CANDIDATES he reviews on /tasks. Nothing here writes to the store, nothing
+// here can create a reminder, and nothing here speaks.
+//
+// Why a separate daemon rather than a loop inside mavend, when extraction has
+// to happen in mavend anyway: the credential. mavpoll set the precedent with the
+// zenmoney token (#125) — the module that talks to a third party holds the
+// secret, reads it from a FILE so it never appears in `ps`, in
+// docker-compose.yml or in shell history, and core never sees it. Core learns
+// that mail exists only as message text on one IPC method; it cannot connect to
+// the mailbox even if it wanted to, and a compromised core yields no mail
+// password.
+//
+// Off unless configured: without -password-file there is nothing to run, and
+// the daemon says so and exits. If core has no `email` block the very first
+// ingest comes back ErrUnknownMethod and this daemon stops polling instead of
+// hammering a socket that will keep refusing.
+//
+// Mail is personal, so the log is counts and UIDs: how many messages were
+// fetched, how many were bulk, how many candidates came back. No subject, no
+// sender, no body, ever — reviewing a candidate is what /tasks is for.
+package main
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"flag"
+	"fmt"
+	"log"
+	"os"
+	"os/signal"
+	"path/filepath"
+	"sort"
+	"strings"
+	"syscall"
+	"time"
+
+	"github.com/kami/maven/internal/email"
+	"github.com/kami/maven/internal/ipc"
+)
+
+func main() {
+	if err := run(os.Args[1:]); err != nil {
+		fmt.Fprintln(os.Stderr, "mavmaild:", err)
+		os.Exit(1)
+	}
+}
+
+func run(args []string) error {
+	fs := flag.NewFlagSet("mavmaild", flag.ContinueOnError)
+	socket := fs.String("socket", "", "core IPC socket path (required)")
+	server := fs.String("imap", "", "IMAP server, host or host:993 (required)")
+	user := fs.String("user", "", "IMAP username (required)")
+	passFile := fs.String("password-file", "", "file holding the IMAP password (required — never passed as a flag value)")
+	mailbox := fs.String("mailbox", "INBOX", "mailbox to read, read-only")
+	interval := fs.Duration("interval", 15*time.Minute, "how often to read the mailbox")
+	lookback := fs.Duration("lookback", 72*time.Hour, "how far back to search on each poll")
+	max := fs.Int("max", 25, "most messages to fetch in one poll")
+	timeout := fs.Duration("timeout", 30*time.Second, "IMAP network timeout")
+	statePath := fs.String("state", "", "file remembering which UIDs were read (default: none — every poll re-reads the window)")
+	if err := fs.Parse(args); err != nil {
+		return err
+	}
+	if *socket == "" {
+		return fmt.Errorf("-socket is required")
+	}
+	if *server == "" || *user == "" || *passFile == "" {
+		return fmt.Errorf("mail reading is off unless configured: set -imap, -user and -password-file")
+	}
+
+	// The password is read from a file, never taken as a flag value: an argv
+	// secret is visible in `ps` to every user on the box and lands in the compose
+	// file and the shell history. Read once at start — a rotated password means a
+	// restart, which is cheaper than re-reading his credential every quarter hour.
+	raw, err := os.ReadFile(*passFile)
+	if err != nil {
+		return fmt.Errorf("read password file: %w", err)
+	}
+	password := strings.TrimSpace(string(raw))
+	if password == "" {
+		return fmt.Errorf("password file %s is empty", *passFile)
+	}
+
+	ctx, stop := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM)
+	defer stop()
+
+	core, err := ipc.DialWait(*socket, 60*time.Second)
+	if err != nil {
+		return err
+	}
+	defer core.Close()
+
+	r := &reader{
+		core:     core,
+		addr:     *server,
+		user:     *user,
+		mailbox:  *mailbox,
+		lookback: *lookback,
+		max:      *max,
+		timeout:  *timeout,
+		state:    newSeenState(*statePath),
+	}
+	if err := r.state.load(); err != nil {
+		// A missing or corrupt state file must not stop mail from being read: the
+		// worst case is re-reading the window, and capture dedupes on text.
+		log.Printf("mavmaild: state: %v (starting from an empty seen-set)", err)
+	}
+
+	// The password is never logged, not even its length.
+	log.Printf("mavmaild: reading %s on %s every %s (lookback %s, max %d/poll)",
+		*mailbox, *server, *interval, *lookback, *max)
+
+	r.pollOnce(ctx, password) // don't idle a full interval on start
+	t := time.NewTicker(*interval)
+	defer t.Stop()
+	for {
+		select {
+		case <-ctx.Done():
+			log.Printf("mavmaild: bye")
+			return nil
+		case <-t.C:
+			if r.disabled {
+				// Core told us mail ingestion is not configured. Nothing will change
+				// without a core restart, and a restart restarts us too.
+				log.Printf("mavmaild: core does not accept mail — idling")
+				return nil
+			}
+			r.pollOnce(ctx, password)
+		}
+	}
+}
+
+// mailIngester — the slice of core this daemon uses. One method: hand over a
+// message. It cannot write a fact, create a reminder or read the store, and the
+// interface says so.
+type mailIngester interface {
+	IngestMail(ctx context.Context, req ipc.IngestMailReq) (ipc.IngestMailResp, error)
+}
+
+type reader struct {
+	core     mailIngester
+	addr     string
+	user     string
+	mailbox  string
+	lookback time.Duration
+	max      int
+	timeout  time.Duration
+	state    *seenState
+
+	// dial — connection seam for the tests; nil ⇒ implicit TLS.
+	dial func(addr string, timeout time.Duration) (*email.Conn, error)
+
+	// disabled — core answered ErrUnknownMethod, i.e. it has no email block.
+	disabled bool
+}
+
+// pollOnce — one read of the mailbox, then one ingest per message.
+//
+// A fetch error aborts this poll and nothing else; the next tick tries again.
+// An ingest error for one message does not skip the rest — one mail the model
+// choked on should not hide the four behind it.
+func (r *reader) pollOnce(ctx context.Context, password string) {
+	msgs, err := r.fetch(password)
+	if err != nil {
+		// The error may name a UID; it never names a subject or a sender.
+		log.Printf("mavmaild: fetch: %v", err)
+		if len(msgs) == 0 {
+			return
+		}
+	}
+	var junk, candidates, created int
+	for _, m := range msgs {
+		if ctx.Err() != nil {
+			return
+		}
+		if m.Junk {
+			junk++
+			// Marked seen without a model call: the header filter already decided,
+			// and re-classifying it every quarter hour would be pure waste.
+			r.state.mark(m.UID)
+			continue
+		}
+		resp, err := r.core.IngestMail(ctx, ipc.IngestMailReq{
+			Mailbox: r.mailbox,
+			UID:     m.UID,
+			From:    m.From,
+			Subject: m.Subject,
+			Date:    m.Date,
+			Body:    m.Body,
+		})
+		if errors.Is(err, ipc.ErrUnknownMethod) {
+			log.Printf("mavmaild: core has no email block configured — mail ingestion is off; stopping")
+			r.disabled = true
+			return
+		}
+		if err != nil {
+			// Not marked seen: an ingest that failed should be retried next poll.
+			log.Printf("mavmaild: ingest uid %d: %v", m.UID, err)
+			continue
+		}
+		r.state.mark(m.UID)
+		candidates += len(resp.TaskIDs)
+		created += resp.Created
+	}
+	if err := r.state.save(); err != nil {
+		log.Printf("mavmaild: state: %v", err)
+	}
+	log.Printf("mavmaild: %s: %d read, %d bulk, %d candidate(s), %d new", r.mailbox, len(msgs), junk, candidates, created)
+}
+
+// fetch reads the mailbox. Messages already in the seen-set are not fetched at
+// all, so a steady mailbox costs one SEARCH per poll and nothing else.
+func (r *reader) fetch(password string) ([]email.Message, error) {
+	f := email.FetchSince{
+		Addr:    r.addr,
+		User:    r.user,
+		Mailbox: r.mailbox,
+		Timeout: r.timeout,
+		Since:   time.Now().Add(-r.lookback),
+		Max:     r.max,
+		Skip:    r.state.seen,
+	}
+	return f.RunWith(password, r.dial)
+}
+
+// ---- seen state ------------------------------------------------------------
+
+// seenState — the UIDs already handed to core, persisted so a restart does not
+// re-read (and re-extract, at multi-second LLM cost) the whole lookback window.
+//
+// Correctness does not depend on it: ipc.CaptureTask dedupes on normalised text
+// among live tasks, so a re-read produces no duplicate rows. This exists to save
+// the model's time, which is why a broken state file is a log line rather than a
+// failure.
+//
+// UIDs are per-mailbox and monotonic, so the set is kept as a high-water mark
+// plus the stragglers above it. If the server ever changes UIDVALIDITY, UIDs
+// reset and the window is simply re-read once — dedupe absorbs it.
+type seenState struct {
+	path  string
+	high  uint32
+	set   map[uint32]bool
+	dirty bool
+}
+
+func newSeenState(path string) *seenState {
+	return &seenState{path: path, set: map[uint32]bool{}}
+}
+
+type seenFile struct {
+	High uint32   `json:"high"`
+	UIDs []uint32 `json:"uids,omitempty"`
+}
+
+func (s *seenState) seen(uid uint32) bool {
+	return uid <= s.high || s.set[uid]
+}
+
+func (s *seenState) mark(uid uint32) {
+	if s.seen(uid) {
+		return
+	}
+	s.set[uid] = true
+	s.dirty = true
+	// Advance the high-water mark through any contiguous run, so the explicit set
+	// stays small on a mailbox read in order.
+	for {
+		next := s.high + 1
+		if !s.set[next] {
+			break
+		}
+		delete(s.set, next)
+		s.high = next
+	}
+}
+
+func (s *seenState) load() error {
+	if s.path == "" {
+		return nil
+	}
+	b, err := os.ReadFile(s.path)
+	if errors.Is(err, os.ErrNotExist) {
+		return nil // first run
+	}
+	if err != nil {
+		return err
+	}
+	var f seenFile
+	if err := json.Unmarshal(b, &f); err != nil {
+		return fmt.Errorf("parse %s: %w", s.path, err)
+	}
+	s.high = f.High
+	for _, u := range f.UIDs {
+		s.set[u] = true
+	}
+	return nil
+}
+
+// save writes the state atomically (temp file + rename), 0600: it is a list of
+// message ids from his mailbox, which is metadata about his mail.
+func (s *seenState) save() error {
+	if s.path == "" || !s.dirty {
+		return nil
+	}
+	uids := make([]uint32, 0, len(s.set))
+	for u := range s.set {
+		uids = append(uids, u)
+	}
+	sort.Slice(uids, func(i, j int) bool { return uids[i] < uids[j] })
+	b, err := json.Marshal(seenFile{High: s.high, UIDs: uids})
+	if err != nil {
+		return err
+	}
+	tmp := s.path + ".tmp"
+	if err := os.MkdirAll(filepath.Dir(s.path), 0o700); err != nil {
+		return err
+	}
+	if err := os.WriteFile(tmp, b, 0o600); err != nil {
+		return err
+	}
+	if err := os.Rename(tmp, s.path); err != nil {
+		return err
+	}
+	s.dirty = false
+	return nil
+}

cmd/mavmaild/main_test.go

@@ -0,0 +1,254 @@
+package main
+
+import (
+	"bufio"
+	"context"
+	"fmt"
+	"net"
+	"os"
+	"path/filepath"
+	"strconv"
+	"strings"
+	"testing"
+	"time"
+
+	"github.com/kami/maven/internal/email"
+	"github.com/kami/maven/internal/ipc"
+)
+
+// ---- a scripted IMAP server, same shape internal/email's tests use ---------
+
+type fakeIMAP struct {
+	msgs map[uint32]string
+	uids []uint32
+	cmds []string
+}
+
+func (f *fakeIMAP) serve(c net.Conn) {
+	defer c.Close()
+	fmt.Fprint(c, "* OK fake ready\r\n")
+	r := bufio.NewReader(c)
+	for {
+		line, err := r.ReadString('\n')
+		if err != nil {
+			return
+		}
+		parts := strings.SplitN(strings.TrimRight(line, "\r\n"), " ", 2)
+		if len(parts) != 2 {
+			return
+		}
+		tag, cmd := parts[0], parts[1]
+		f.cmds = append(f.cmds, cmd)
+		upper := strings.ToUpper(cmd)
+		switch {
+		case strings.HasPrefix(upper, "LOGIN"), strings.HasPrefix(upper, "EXAMINE"):
+			fmt.Fprintf(c, "%s OK\r\n", tag)
+		case strings.HasPrefix(upper, "UID SEARCH"):
+			var ids []string
+			for _, u := range f.uids {
+				ids = append(ids, strconv.FormatUint(uint64(u), 10))
+			}
+			fmt.Fprintf(c, "* SEARCH %s\r\n%s OK\r\n", strings.Join(ids, " "), tag)
+		case strings.HasPrefix(upper, "UID FETCH"):
+			uid, _ := strconv.ParseUint(strings.Fields(cmd)[2], 10, 32)
+			if raw, ok := f.msgs[uint32(uid)]; ok {
+				fmt.Fprintf(c, "* 1 FETCH (UID %d BODY[] {%d}\r\n%s)\r\n", uid, len(raw), raw)
+			}
+			fmt.Fprintf(c, "%s OK\r\n", tag)
+		case strings.HasPrefix(upper, "LOGOUT"):
+			fmt.Fprintf(c, "* BYE\r\n%s OK\r\n", tag)
+			return
+		default:
+			fmt.Fprintf(c, "%s BAD\r\n", tag)
+		}
+	}
+}
+
+func (f *fakeIMAP) dial(_ string, timeout time.Duration) (*email.Conn, error) {
+	cli, srv := net.Pipe()
+	go f.serve(srv)
+	return email.NewConn(cli, timeout)
+}
+
+// ---- a fake core -----------------------------------------------------------
+
+type fakeCore struct {
+	got  []ipc.IngestMailReq
+	resp ipc.IngestMailResp
+	err  error
+}
+
+func (c *fakeCore) IngestMail(_ context.Context, req ipc.IngestMailReq) (ipc.IngestMailResp, error) {
+	c.got = append(c.got, req)
+	if c.err != nil {
+		return ipc.IngestMailResp{}, c.err
+	}
+	return c.resp, nil
+}
+
+func mail(subject, body string, extraHeaders ...string) string {
+	h := "Subject: " + subject + "\r\nFrom: a@b.c\r\nContent-Type: text/plain; charset=utf-8\r\n"
+	for _, e := range extraHeaders {
+		h += e + "\r\n"
+	}
+	return h + "\r\n" + body + "\r\n"
+}
+
+func newTestReader(t *testing.T, f *fakeIMAP, core *fakeCore, statePath string) *reader {
+	t.Helper()
+	return &reader{
+		core: core, addr: "mail.example:993", user: "kami", mailbox: "INBOX",
+		lookback: 72 * time.Hour, max: 25, timeout: 5 * time.Second,
+		state: newSeenState(statePath),
+		dial:  f.dial,
+	}
+}
+
+func TestPollHandsMessagesToCore(t *testing.T) {
+	f := &fakeIMAP{
+		uids: []uint32{1, 2},
+		msgs: map[uint32]string{
+			1: mail("Счёт", "Оплатить до 5 августа."),
+			2: mail("Скидки", "Sale!", "List-Unsubscribe: <mailto:u@x>"),
+		},
+	}
+	core := &fakeCore{resp: ipc.IngestMailResp{TaskIDs: []int64{1}, Created: 1}}
+	r := newTestReader(t, f, core, "")
+	r.pollOnce(context.Background(), "secret")
+
+	// The newsletter is filtered before core is asked: only the real mail crosses.
+	if len(core.got) != 1 {
+		t.Fatalf("core saw %d messages, want 1 (the bulk one must not cross): %+v", len(core.got), core.got)
+	}
+	got := core.got[0]
+	if got.UID != 1 || got.Mailbox != "INBOX" || got.Subject != "Счёт" {
+		t.Errorf("ingest req = %+v", got)
+	}
+	if !strings.Contains(got.Body, "Оплатить") {
+		t.Errorf("body = %q", got.Body)
+	}
+}
+
+// A second poll must not re-send what core already saw — extraction is a
+// multi-second LLM call per message.
+func TestPollSkipsSeenUIDs(t *testing.T) {
+	f := &fakeIMAP{uids: []uint32{5}, msgs: map[uint32]string{5: mail("Счёт", "текст")}}
+	core := &fakeCore{}
+	r := newTestReader(t, f, core, "")
+	r.pollOnce(context.Background(), "secret")
+	r.pollOnce(context.Background(), "secret")
+	if len(core.got) != 1 {
+		t.Errorf("core saw %d messages over two polls, want 1", len(core.got))
+	}
+}
+
+// An ingest that failed is NOT marked seen: the next poll retries it.
+func TestPollRetriesFailedIngest(t *testing.T) {
+	f := &fakeIMAP{uids: []uint32{5}, msgs: map[uint32]string{5: mail("Счёт", "текст")}}
+	core := &fakeCore{err: fmt.Errorf("llama-server is warming up")}
+	r := newTestReader(t, f, core, "")
+	r.pollOnce(context.Background(), "secret")
+	core.err = nil
+	r.pollOnce(context.Background(), "secret")
+	if len(core.got) != 2 {
+		t.Errorf("core saw %d attempts, want 2 (a failed ingest is retried)", len(core.got))
+	}
+}
+
+// Core without an email block ⇒ stop, don't hammer the socket.
+func TestPollStopsWhenCoreRefusesMail(t *testing.T) {
+	f := &fakeIMAP{uids: []uint32{1, 2}, msgs: map[uint32]string{1: mail("a", "b"), 2: mail("c", "d")}}
+	core := &fakeCore{err: fmt.Errorf("call: %w", ipc.ErrUnknownMethod)}
+	r := newTestReader(t, f, core, "")
+	r.pollOnce(context.Background(), "secret")
+	if !r.disabled {
+		t.Error("ErrUnknownMethod must disable the reader")
+	}
+	if len(core.got) != 1 {
+		t.Errorf("core saw %d messages, want 1 — stop at the first refusal", len(core.got))
+	}
+}
+
+func TestSeenStatePersists(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "state", "seen.json")
+	f := &fakeIMAP{uids: []uint32{9}, msgs: map[uint32]string{9: mail("Счёт", "текст")}}
+	core := &fakeCore{}
+	r := newTestReader(t, f, core, path)
+	r.pollOnce(context.Background(), "secret")
+
+	fi, err := os.Stat(path)
+	if err != nil {
+		t.Fatalf("state file: %v", err)
+	}
+	// A list of message ids from his mailbox is metadata about his mail.
+	if perm := fi.Mode().Perm(); perm != 0o600 {
+		t.Errorf("state file mode = %v, want 0600", perm)
+	}
+
+	// A fresh reader with the same state file must not re-read the message.
+	core2 := &fakeCore{}
+	r2 := newTestReader(t, f, core2, path)
+	if err := r2.state.load(); err != nil {
+		t.Fatalf("load: %v", err)
+	}
+	r2.pollOnce(context.Background(), "secret")
+	if len(core2.got) != 0 {
+		t.Errorf("after a restart core saw %d messages, want 0", len(core2.got))
+	}
+}
+
+func TestSeenStateHighWaterMark(t *testing.T) {
+	s := newSeenState("")
+	s.mark(1)
+	s.mark(3)
+	s.mark(2)
+	if s.high != 3 {
+		t.Errorf("high = %d, want 3 (contiguous run collapses)", s.high)
+	}
+	if len(s.set) != 0 {
+		t.Errorf("explicit set = %v, want empty", s.set)
+	}
+	if !s.seen(2) || s.seen(4) {
+		t.Errorf("seen(2)=%v seen(4)=%v", s.seen(2), s.seen(4))
+	}
+}
+
+func TestSeenStateCorruptFileIsNotFatal(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "seen.json")
+	if err := os.WriteFile(path, []byte("{not json"), 0o600); err != nil {
+		t.Fatal(err)
+	}
+	s := newSeenState(path)
+	if err := s.load(); err == nil {
+		t.Error("a corrupt state file should report an error the caller logs")
+	}
+	if s.seen(1) {
+		t.Error("a corrupt state file must leave an empty seen-set, not a poisoned one")
+	}
+}
+
+// Off unless configured, and the credential is never a flag value.
+func TestRunRequiresConfig(t *testing.T) {
+	if err := run([]string{}); err == nil {
+		t.Error("no -socket must be an error")
+	}
+	if err := run([]string{"-socket", "/tmp/nope.sock"}); err == nil {
+		t.Error("no mailbox configuration must be an error, not a default mailbox")
+	}
+	// There is no -password flag at all: only -password-file.
+	if err := run([]string{"-socket", "/x", "-imap", "h", "-user", "u", "-password", "p"}); err == nil ||
+		!strings.Contains(err.Error(), "flag provided but not defined") {
+		t.Errorf("a -password flag must not exist; err = %v", err)
+	}
+}
+
+func TestRunRejectsEmptyPasswordFile(t *testing.T) {
+	path := filepath.Join(t.TempDir(), "pass")
+	if err := os.WriteFile(path, []byte("  \n"), 0o600); err != nil {
+		t.Fatal(err)
+	}
+	err := run([]string{"-socket", "/x/y.sock", "-imap", "h", "-user", "u", "-password-file", path})
+	if err == nil || !strings.Contains(err.Error(), "empty") {
+		t.Errorf("an empty password file must be refused before dialling; err = %v", err)
+	}
+}

cmd/mavpoll/main.go

@@ -9,6 +9,12 @@
 // Two sources, each its own provenance (the loop's rules trust source):
 //   - netdata  → poll:netdata      resource alarms (disk/mem/cert/temp)
 //   - kuma     → poll:uptimekuma   service up/down (the source of truth for it)
+//   - zenmoney → poll:zenmoney     spending/income totals (Vikunja #125)
+//
+// The zenmoney source is why the token lives HERE and not in core: the poller
+// already owns every other third-party credential, it holds no store key, and
+// core never needs to know an account exists to answer a question about a fact
+// the poller wrote. It is off unless -zenmoney-token-file is given.
 //
 // Netdata needs no auth over the wg-fronted net. Kuma's /metrics needs an API
 // key (basic-auth); without -kuma the whole kuma path is skipped (netdata-only
@@ -37,6 +43,7 @@ import (
 	"time"
 
 	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/zenmoney"
 )
 
 func main() {
@@ -52,6 +59,9 @@ func run(args []string) error {
 	netdataURL := fs.String("netdata", "http://127.0.0.1:19999", "netdata base URL ('' to disable)")
 	kumaURL := fs.String("kuma", "", "uptime-kuma metrics URL, e.g. http://127.0.0.1:3001/metrics ('' to disable)")
 	kumaKey := fs.String("kuma-key", "", "uptime-kuma API key (basic-auth username)")
+	zenTokenFile := fs.String("zenmoney-token-file", "", "file holding the zenmoney API token ('' disables money tracking)")
+	zenURL := fs.String("zenmoney-url", zenmoney.DefaultBaseURL, "zenmoney API base URL (tests/self-hosted proxies)")
+	zenInterval := fs.Duration("zenmoney-interval", time.Hour, "how often to read zenmoney (money does not move every minute)")
 	wgIface := fs.String("wg", "", "wireguard interface for the presence signal, e.g. wg0 or 'all' ('' to disable)")
 	wgCmd := fs.String("wg-cmd", "wg", "wg binary (use e.g. 'sudo wg' if the poller lacks CAP_NET_ADMIN)")
 	interval := fs.Duration("interval", 60*time.Second, "poll cadence")
@@ -62,8 +72,24 @@ func run(args []string) error {
 	if *socket == "" {
 		return fmt.Errorf("-socket is required")
 	}
-	if *netdataURL == "" && *kumaURL == "" && *wgIface == "" {
-		return fmt.Errorf("nothing to poll: set -netdata, -kuma and/or -wg")
+	if *netdataURL == "" && *kumaURL == "" && *wgIface == "" && *zenTokenFile == "" {
+		return fmt.Errorf("nothing to poll: set -netdata, -kuma, -wg and/or -zenmoney-token-file")
+	}
+
+	// The token is read from a file, never taken as a flag value: an argv token
+	// is visible in `ps` to every user on the box and lands in the compose file
+	// and the shell history. Read once at start — a rotated token means a
+	// restart, which is cheaper than re-reading his credential every hour.
+	var zen *zenmoney.Client
+	if *zenTokenFile != "" {
+		raw, err := os.ReadFile(*zenTokenFile)
+		if err != nil {
+			return fmt.Errorf("read zenmoney token: %w", err)
+		}
+		zen, err = zenmoney.New(strings.TrimSpace(string(raw)), *zenURL, *timeout*3)
+		if err != nil {
+			return err
+		}
 	}
 
 	ctx, stop := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM)
@@ -83,9 +109,13 @@ func run(args []string) error {
 		kumaKey:    *kumaKey,
 		wgIface:    *wgIface,
 		wgCmd:      *wgCmd,
+		zen:        zen,
+		zenEvery:   *zenInterval,
 	}
 
-	log.Printf("mavpoll: polling every %s (netdata=%q kuma=%q wg=%q)", *interval, *netdataURL, *kumaURL, *wgIface)
+	// The token is never logged, not even its length.
+	log.Printf("mavpoll: polling every %s (netdata=%q kuma=%q wg=%q zenmoney=%v every %s)",
+		*interval, *netdataURL, *kumaURL, *wgIface, zen != nil, *zenInterval)
 	p.pollOnce(ctx) // fire immediately; don't idle a full interval on start
 	t := time.NewTicker(*interval)
 	defer t.Stop()
@@ -108,6 +138,12 @@ type poller struct {
 	kumaKey    string
 	wgIface    string
 	wgCmd      string
+
+	// zen is nil unless a token file was configured — money tracking is a
+	// capability, off by default like weather and telegram.
+	zen      *zenmoney.Client
+	zenEvery time.Duration
+	zenLast  time.Time
 }
 
 // pollOnce — one sweep of both sources. A failure in one source logs and does
@@ -129,6 +165,66 @@ func (p *poller) pollOnce(ctx context.Context) {
 			log.Printf("mavpoll: wg: %v", err)
 		}
 	}
+	// Money on its own, much slower cadence: a bank feed that updates hourly
+	// polled every minute is 60 pointless reads of his financial history.
+	if p.zen != nil && now.Sub(p.zenLast) >= p.zenEvery {
+		p.zenLast = now
+		if err := p.pollZenmoney(ctx, now); err != nil {
+			log.Printf("mavpoll: zenmoney: %v", err)
+		}
+	}
+}
+
+// ---- zenmoney: spending/income totals → money facts ------------------------
+
+// pollZenmoney reads today's and this month's totals and writes them as
+// facts(kind=env, source=poll:zenmoney) (Vikunja #125).
+//
+// Two properties this function exists to hold:
+//
+//   - An empty or failed read writes NOTHING. zenmoney.Summary.Value() refuses
+//     to encode a summary built from zero transactions, so a poller that cannot
+//     reach the API leaves the last good fact in place rather than overwriting
+//     it with a zero Maven would then recite as fact.
+//   - Nothing about the money leaves the box except the diff request itself, to
+//     the service that already holds his bank sessions. The totals are written
+//     to the store and read back only when he asks; they are never search input
+//     and no tick rule fires on them.
+//
+// Both windows are read from one diff call each. Two calls an hour against an
+// API whose whole job is this is not worth caching.
+// moneyWindow — one fact key and the period it covers.
+type moneyWindow struct {
+	key      string
+	from, to time.Time
+}
+
+func (p *poller) pollZenmoney(ctx context.Context, now time.Time) error {
+	dFrom, dTo := zenmoney.DayWindow(now)
+	mFrom, mTo := zenmoney.MonthWindow(now)
+	windows := []moneyWindow{
+		{zenmoney.KeySpentToday, dFrom, dTo},
+		{zenmoney.KeySpentMonth, mFrom, mTo},
+	}
+	var firstErr error
+	for _, w := range windows {
+		sum, err := p.zen.Since(ctx, w.from, w.to)
+		if err != nil {
+			if firstErr == nil {
+				firstErr = err
+			}
+			continue
+		}
+		val, ok := sum.Value()
+		if !ok {
+			// Nothing read. Silence, not a zero.
+			continue
+		}
+		if err := p.writeIfChangedRaw(ctx, w.key, zenmoney.Source, val, now); err != nil && firstErr == nil {
+			firstErr = err
+		}
+	}
+	return firstErr
 }
 
 // ---- wireguard: latest handshake → presence signal -------------------------
@@ -301,6 +397,31 @@ func (p *poller) writeIfChanged(ctx context.Context, key, source, val string, no
 	return nil
 }
 
+// writeIfChangedRaw is writeIfChanged for values that are already JSON (the
+// money facts store an object, not a string). Kept separate rather than
+// generalising writeIfChanged, because the string-valued env facts encoding
+// their own value is the convention the rules rely on.
+//
+// The log line names the key and the source, never the figures: mavpoll's log
+// is not the place his spending ends up.
+func (p *poller) writeIfChangedRaw(ctx context.Context, key, source, jsonVal string, now time.Time) error {
+	prev, err := p.core.LatestFactBySource(ctx, key, source)
+	switch {
+	case err == nil && prev.Value == jsonVal:
+		return nil
+	case err != nil && err != ipc.ErrNoFact && !isNoFact(err):
+		return fmt.Errorf("read %s: %w", key, err)
+	}
+	if _, err := p.core.WriteFact(ctx, ipc.WriteFactReq{
+		Ts: now, Kind: "env", Key: key, Value: jsonVal,
+		Source: source, Confidence: 1.0,
+	}); err != nil {
+		return fmt.Errorf("write %s: %w", key, err)
+	}
+	log.Printf("mavpoll: %s updated (%s)", key, source)
+	return nil
+}
+
 // isNoFact — ErrNoFact rehydrated over the wire is wrapped (fmt.Errorf %w), so
 // errors.Is is the right check; keep a helper so the switch above reads clean.
 func isNoFact(err error) bool {

cmd/mavpoll/main_test.go

@@ -1,8 +1,17 @@
 package main
 
 import (
+	"context"
 	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"os"
+	"strings"
 	"testing"
+	"time"
+
+	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/zenmoney"
 )
 
 func TestMaxSeverity(t *testing.T) {
@@ -62,3 +71,120 @@ func TestParseMaxHandshake(t *testing.T) {
 		}
 	}
 }
+
+// ---- zenmoney (Vikunja #125) ----------------------------------------------
+
+// factCore records the facts the poller wrote and answers "no fact yet".
+type factCore struct {
+	ipc.UnimplementedCoreAPI
+
+	written []ipc.WriteFactReq
+	prev    map[string]string
+}
+
+func (c *factCore) LatestFactBySource(_ context.Context, key, source string) (ipc.Fact, error) {
+	if v, ok := c.prev[key+"|"+source]; ok {
+		return ipc.Fact{Key: key, Source: source, Value: v}, nil
+	}
+	return ipc.Fact{}, ipc.ErrNoFact
+}
+
+func (c *factCore) WriteFact(_ context.Context, req ipc.WriteFactReq) (int64, error) {
+	c.written = append(c.written, req)
+	return int64(len(c.written)), nil
+}
+
+func zenFixtureServer(t *testing.T, body []byte, status int) *httptest.Server {
+	t.Helper()
+	return httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		if status != http.StatusOK {
+			w.WriteHeader(status)
+			return
+		}
+		w.Write(body)
+	}))
+}
+
+func TestPollZenmoneyWritesMoneyFacts(t *testing.T) {
+	body, err := os.ReadFile("../../internal/zenmoney/testdata/diff.json")
+	if err != nil {
+		t.Fatal(err)
+	}
+	srv := zenFixtureServer(t, body, http.StatusOK)
+	defer srv.Close()
+	zen, err := zenmoney.New("tok", srv.URL, time.Second)
+	if err != nil {
+		t.Fatal(err)
+	}
+	core := &factCore{}
+	p := &poller{core: core, zen: zen}
+	now := time.Date(2026, 8, 1, 21, 0, 0, 0, time.UTC)
+	if err := p.pollZenmoney(context.Background(), now); err != nil {
+		t.Fatal(err)
+	}
+	if len(core.written) != 2 {
+		t.Fatalf("wrote %d facts, want today + month", len(core.written))
+	}
+	for _, f := range core.written {
+		if f.Kind != "env" || f.Source != zenmoney.Source {
+			t.Errorf("fact = %+v, want kind=env source=%s", f, zenmoney.Source)
+		}
+		if _, err := zenmoney.ParseFactValue(f.Value); err != nil {
+			t.Errorf("fact value %q does not decode: %v", f.Value, err)
+		}
+	}
+}
+
+// A read that returns nothing for the window writes NOTHING. Silence, not a
+// zero: an invented 0 would be recited back to him as fact.
+func TestPollZenmoneyWritesNothingWhenEmpty(t *testing.T) {
+	srv := zenFixtureServer(t, []byte(`{"serverTimestamp":1,"instrument":[],"transaction":[]}`), http.StatusOK)
+	defer srv.Close()
+	zen, _ := zenmoney.New("tok", srv.URL, time.Second)
+	core := &factCore{}
+	p := &poller{core: core, zen: zen}
+	if err := p.pollZenmoney(context.Background(), time.Now()); err != nil {
+		t.Fatal(err)
+	}
+	if len(core.written) != 0 {
+		t.Errorf("wrote %+v, want no fact at all", core.written)
+	}
+}
+
+// An API failure must not overwrite the last good total either.
+func TestPollZenmoneyFailureWritesNothing(t *testing.T) {
+	srv := zenFixtureServer(t, nil, http.StatusUnauthorized)
+	defer srv.Close()
+	zen, _ := zenmoney.New("bad", srv.URL, time.Second)
+	core := &factCore{}
+	p := &poller{core: core, zen: zen}
+	if err := p.pollZenmoney(context.Background(), time.Now()); err == nil {
+		t.Error("want the 401 reported")
+	}
+	if len(core.written) != 0 {
+		t.Errorf("wrote %+v on a failed read", core.written)
+	}
+}
+
+// Unchanged totals do not churn the facts table.
+func TestWriteIfChangedRawSkipsUnchanged(t *testing.T) {
+	core := &factCore{prev: map[string]string{
+		zenmoney.KeySpentToday + "|" + zenmoney.Source: `{"count":1}`,
+	}}
+	p := &poller{core: core}
+	if err := p.writeIfChangedRaw(context.Background(), zenmoney.KeySpentToday, zenmoney.Source, `{"count":1}`, time.Now()); err != nil {
+		t.Fatal(err)
+	}
+	if len(core.written) != 0 {
+		t.Errorf("wrote %+v for an unchanged value", core.written)
+	}
+}
+
+// Money tracking is off unless configured: no token file, no zenmoney client,
+// and the poller still refuses to start with nothing at all to poll.
+func TestRunRequiresSomethingToPoll(t *testing.T) {
+	err := run([]string{"-socket", "/tmp/nope.sock", "-netdata", "", "-kuma", "", "-wg", ""})
+	if err == nil || !strings.Contains(err.Error(), "nothing to poll") {
+		t.Errorf("err = %v, want a 'nothing to poll' refusal", err)
+	}
+}

cmd/mavweb/main.go

@@ -25,6 +25,7 @@ import (
 	"github.com/kami/maven/internal/audio"
 	"github.com/kami/maven/internal/ipc"
 	"github.com/kami/maven/internal/pattern"
+	"github.com/kami/maven/internal/tasks"
 	"github.com/kami/maven/internal/voice"
 	"github.com/kami/maven/internal/webauthn"
 )
@@ -840,6 +841,10 @@ type taskRow struct {
 	Due      string
 	Created  string
 	Resolved string
+	// Why — the ranker's reason for this row's position (Vikunja #129), in
+	// Russian, empty when nothing distinguished the task. Blank is the honest
+	// rendering: he never said this one mattered more.
+	Why string
 }
 
 // handleTasks serves the task review surface (GET) and the four writes it
@@ -878,20 +883,46 @@ func handleTasks(w http.ResponseWriter, r *http.Request, core ipc.CoreAPI) {
 		http.Error(w, "tasks error: "+err.Error(), http.StatusBadGateway)
 		return
 	}
-	var cands, open, resolved []taskRow
+	// Live rows are ordered by the same ranker the spoken list uses, so the page
+	// and the voice reply can never disagree about what comes first. Resolved
+	// rows keep store order (newest first) — ranking finished work is pointless.
+	var live []tasks.Item
+	var resolved []taskRow
 	for _, t := range all {
+		switch t.Status {
+		case "candidate", "open":
+			live = append(live, tasks.Item{
+				ID: t.ID, Text: t.Text, Status: t.Status,
+				Created: t.CreatedTs, Due: t.Due, Weight: t.Weight,
+			})
+		default:
+			resolved = append(resolved, taskRow{
+				ID: t.ID, Text: t.Text, Source: t.Source, Evidence: t.Evidence,
+				Status: t.Status, Created: fmtTaskTime(&t.CreatedTs),
+				Due: fmtTaskDate(t.Due), Resolved: fmtTaskTime(t.Resolved),
+			})
+		}
+	}
+	byID := make(map[int64]ipc.Task, len(all))
+	for _, t := range all {
+		byID[t.ID] = t
+	}
+	var cands, open []taskRow
+	for _, r := range tasks.Rank(live, time.Now()) {
+		t := byID[r.ID]
 		row := taskRow{
 			ID: t.ID, Text: t.Text, Source: t.Source, Evidence: t.Evidence,
 			Status: t.Status, Created: fmtTaskTime(&t.CreatedTs),
 			Due: fmtTaskDate(t.Due), Resolved: fmtTaskTime(t.Resolved),
+			Why: r.Reason,
 		}
-		switch t.Status {
-		case "candidate":
+		if t.Status == "candidate" {
+			// A candidate's due date is Maven's reading of a mail, so its
+			// ranking reason is not shown as if he had set a priority.
+			row.Why = ""
 			cands = append(cands, row)
-		case "open":
+		} else {
 			open = append(open, row)
-		default:
-			resolved = append(resolved, row)
 		}
 	}
 	w.Header().Set("Content-Type", "text/html; charset=utf-8")
@@ -916,6 +947,18 @@ func applyTaskPost(ctx context.Context, core ipc.CoreAPI, r *http.Request) (stri
 			return "", errors.New("empty task text")
 		}
 		req := ipc.CaptureTaskReq{Text: text, Source: "tap:web", Status: "open", Ts: time.Now()}
+		// Importance is his, stated on the form. Out-of-range values are
+		// clamped rather than rejected — a bad select is not worth a 400.
+		if v := r.FormValue("weight"); v != "" {
+			var wgt int
+			if n, _ := fmt.Sscanf(v, "%d", &wgt); n != 1 || wgt < 0 {
+				return "", fmt.Errorf("bad weight %q", v)
+			}
+			if wgt > tasks.MaxWeight {
+				wgt = tasks.MaxWeight
+			}
+			req.Weight = wgt
+		}
 		if d := r.FormValue("due"); d != "" {
 			due, err := time.ParseInLocation("2006-01-02", d, time.Local)
 			if err != nil {

cmd/mavweb/tasks.html

@@ -9,6 +9,11 @@
 <input type=hidden name=action value=add>
 <input type=text name=text placeholder="что нужно сделать" size=44 required>
 <input type=date name=due title="due date (optional)">
+<select name=weight title="importance (optional)">
+<option value=0>normal</option>
+<option value=2>важно</option>
+<option value=3>срочно</option>
+</select>
 <button class=btn>add</button>
 </form>
 </section>
@@ -38,10 +43,12 @@
 
 <section class=card>
 <h2 class=card-title>open <span class=badge>{{len .Open}}</span></h2>
+<div class=hint>most pressing first — by the deadlines and the urgency you gave, nothing guessed.</div>
 {{if .Open}}<div class=scroll><table>
-<tr><th>task</th><th>from</th><th>due</th><th>captured</th><th></th><th></th></tr>
+<tr><th>task</th><th>why</th><th>from</th><th>due</th><th>captured</th><th></th><th></th></tr>
 {{range .Open}}<tr>
 <td class=text-max>{{.Text}}</td>
+<td class=hint>{{.Why}}</td>
 <td class=hint>{{.Source}}</td>
 <td>{{.Due}}</td>
 <td class=muted>{{.Created}}</td>

cmd/mavweb/tasks_test.go

@@ -10,6 +10,7 @@ import (
 	"time"
 
 	"github.com/kami/maven/internal/ipc"
+	"github.com/kami/maven/internal/tasks"
 )
 
 // fakeTaskCore serves the /tasks handler: a canned list plus a log of the
@@ -164,3 +165,61 @@ func TestHandleTasksNoCore(t *testing.T) {
 		t.Errorf("status = %d, want 503", rec.Code)
 	}
 }
+
+// The open list is ordered by the ranker, and the reason is shown so the page
+// says why a task is first instead of asking him to trust the order.
+func TestHandleTasksOrdersOpenByRank(t *testing.T) {
+	now := time.Now()
+	due := now
+	core := &fakeTaskCore{tasks: []ipc.Task{
+		{ID: 1, Text: "купить молоко", Status: "open", CreatedTs: now},
+		{ID: 2, Text: "оплатить интернет", Status: "open", CreatedTs: now, Due: &due},
+	}}
+	rec := httptest.NewRecorder()
+	handleTasks(rec, httptest.NewRequest(http.MethodGet, "/tasks", nil), core)
+	body := rec.Body.String()
+	if strings.Index(body, "оплатить интернет") > strings.Index(body, "купить молоко") {
+		t.Error("want the dated task rendered first")
+	}
+	if !strings.Contains(body, "сегодня") {
+		t.Error("want the ranker's reason shown in the why column")
+	}
+}
+
+// A candidate is ranked into place but never carries a priority reason: its due
+// date is Maven's reading of a mail, not something he stated.
+func TestHandleTasksHidesCandidateReason(t *testing.T) {
+	now := time.Now()
+	due := now
+	core := &fakeTaskCore{tasks: []ipc.Task{
+		{ID: 1, Text: "продлить страховку", Status: "candidate", CreatedTs: now, Due: &due},
+	}}
+	rec := httptest.NewRecorder()
+	handleTasks(rec, httptest.NewRequest(http.MethodGet, "/tasks", nil), core)
+	if strings.Contains(rec.Body.String(), "сегодня") {
+		t.Error("a candidate must not be shown with a priority reason")
+	}
+}
+
+func TestApplyTaskPostCarriesWeight(t *testing.T) {
+	core := &fakeTaskCore{created: true}
+	form := url.Values{"action": {"add"}, "text": {"оплатить интернет"}, "weight": {"3"}}
+	req := httptest.NewRequest(http.MethodPost, "/tasks", strings.NewReader(form.Encode()))
+	req.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	handleTasks(httptest.NewRecorder(), req, core)
+	if len(core.captured) != 1 || core.captured[0].Weight != 3 {
+		t.Fatalf("captured = %+v, want weight 3", core.captured)
+	}
+}
+
+// Out of range clamps rather than 400s; a non-number is a real client error.
+func TestApplyTaskPostClampsWeight(t *testing.T) {
+	core := &fakeTaskCore{created: true}
+	form := url.Values{"action": {"add"}, "text": {"что-то"}, "weight": {"99"}}
+	req := httptest.NewRequest(http.MethodPost, "/tasks", strings.NewReader(form.Encode()))
+	req.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	handleTasks(httptest.NewRecorder(), req, core)
+	if core.captured[0].Weight != tasks.MaxWeight {
+		t.Errorf("weight = %d, want the cap", core.captured[0].Weight)
+	}
+}

deploy/README.md

@@ -36,6 +36,67 @@ present (see `.dockerignore`).
 | `/var/lib/maven` (volume)  | encrypted db at rest                    |
 | `/dev/shm` (tmpfs)         | decrypted db working copy (RAM only)    |
 
+## Reading the outside world (off by default)
+
+`mavend.json` ships without a `feeds` block, which means no RSS/Atom feed is
+fetched and no outbound request is made. Switching it on is adding the block:
+
+```json
+"feeds": {
+  "poll_interval": "30m",
+  "max_items": 5,
+  "max_age": "24h",
+  "sources": [
+    { "name": "habr", "url": "https://habr.com/ru/rss/best/daily/",
+      "category": "технологии", "exclude": ["реклама"] }
+  ]
+}
+```
+
+What it does and does not do:
+
+- items are written as notes with source `rss:<name>`, visible on `/dash`;
+- **nothing is announced.** She reads them back when asked — "что нового в
+  лентах?", "что нового по технологиям?" — and never on arrival. There is no
+  severity or channel knob here on purpose;
+- the fetcher is allowlisted to the hosts of the configured feeds, plus any
+  `allow_hosts`. It refuses non-http(s) schemes and every private address
+  (loopback, the LAN, the `10.42.0.0/24` wg range, cloud metadata). It caps the
+  response at 2 MiB and redirects at 3, and makes at most one request per host
+  per second. See `internal/webfetch`;
+- how far each feed was read is stored as a config fact `rss:latest:<name>`, so
+  a restart does not re-note yesterday's headlines.
+
+### Reading a page (`crawl`, also off by default)
+
+There is no `crawl` block either, so no page is fetched. Two halves, separately
+switched:
+
+```json
+"crawl": {
+  "on_demand": true,
+  "interval": "6h",
+  "max_runes": 4000,
+  "watches": [
+    { "name": "changelog", "url": "https://example.org/changelog", "interval": "12h" }
+  ]
+}
+```
+
+- `on_demand` lets her read a page he names in the utterance: "посмотри
+  https://example.org/x — что там?". The page becomes context for his question,
+  and only the URL leaves the box. Without a URL nothing is fetched, so this is
+  a fallback and not a habit;
+- `watches` re-reads a fixed list on its interval and writes a note when the
+  text changed. Like the feeds, it announces nothing;
+- the answer path sits **last** in the query chain, behind his memory, his notes
+  and (once wired) the local Kiwix ZIMs. A local read costs nothing;
+- `robots.txt` is fetched first and obeyed with no override; a `Disallow` is a
+  refusal she says out loud. `Crawl-delay` is honoured;
+- same guarded fetcher as the feeds: allowlist/denylist, no private addresses,
+  size cap, redirect cap, timeout, one request per host per second;
+- dedup state is the config fact `crawl:hash:<name>`.
+
 ## Not yet verified / host-dependent
 
 This stack is correct-by-construction but has **not been build-tested here**

docker-compose.yml

@@ -101,9 +101,42 @@ services:
               "-netdata", "http://127.0.0.1:19999",
               "-kuma", "http://127.0.0.1:3001/metrics",
               "-kuma-key", "uk5_mavpoll-key"]
+              # Money tracking (Vikunja #125) is OFF: it needs a zenmoney token,
+              # which mavpoll reads from a FILE so it never appears in `ps`, in
+              # this file, or in shell history. To enable, mount the token and
+              # append: "-zenmoney-token-file", "/run/secrets/zenmoney.token"
+              # (optionally "-zenmoney-interval", "1h"). Core never sees the
+              # token — the poller writes facts(kind=env, source=poll:zenmoney)
+              # and mavend only reads those back when he asks.
     depends_on: [mavend]
     volumes:
       - sockets:/run/maven
+      # - ./deploy/zenmoney.token:/run/secrets/zenmoney.token:ro
+
+  # The mail reader (Vikunja #246) is OFF and commented out: it needs an IMAP
+  # account, and there is none on this box. mavmaild reads the password from a
+  # FILE so it never appears in `ps`, in this file, or in shell history — the
+  # same rule mavpoll follows for the zenmoney token. Core never sees the
+  # password: the reader hands core message text on one IPC method, and core
+  # writes what the model extracts as task CANDIDATES he reviews on /tasks.
+  # Nothing here can create a reminder, so a misread mail cannot fire.
+  #
+  # To enable: write the password to deploy/imap.password (0600, gitignored),
+  # add an "email": {} block to deploy/mavend.json, and uncomment this service.
+  # mavmaild:
+  #   <<: *image
+  #   command: ["mavmaild", "-socket", "/run/maven/mavend.sock",
+  #             "-imap", "imap.example.org:993",
+  #             "-user", "kami@example.org",
+  #             "-password-file", "/run/secrets/imap.password",
+  #             "-mailbox", "INBOX",
+  #             "-interval", "15m",
+  #             "-state", "/var/lib/maven/mail-seen.json"]
+  #   depends_on: [mavend]
+  #   volumes:
+  #     - sockets:/run/maven
+  #     - dbdata:/var/lib/maven
+  #     - ./deploy/imap.password:/run/secrets/imap.password:ro
 
 volumes:
   dbdata:

docs/plans/13-rss-news-feeds.md

@@ -27,3 +27,28 @@
 7. Add voice query handler — `"что нового?"` queries `RecentNotes` filtered by source prefix `rss:` and phrases via `phraser.PhraseQuery`
 8. Add `feeds` block to `config.Config` and `deploy/mavend.json`
 9. Test with a live RSS feed (e.g., `https://news.ycombinator.com/rss`) — verify items appear in notes table
+
+---
+
+## Shipped 2026-08-01 (#258)
+
+`internal/webfetch` (the guarded HTTP door: scheme, allow/deny hosts, private-address
+refusal in the dialer, size cap, redirect cap, per-host rate limit), `internal/rss`
+(RSS 2.0 + Atom parser, poller with durable marks and a keyword filter),
+`cmd/mavend/feeds.go` (ticker, fetcher adapter, `rss:latest:<feed>` fact marks),
+config block `feeds`, and the `feeds` query source with `router.ParseFeedQuery`.
+
+Deviations from the plan above, both deliberate:
+
+- **Step 5 (breaking-news nudges) was not built.** A feed that dispatches is a nag,
+  and the one thing Maven is not is a nag. Items are read when asked and nowhere else.
+  If breaking news is ever wanted, it belongs behind the existing delivery policy
+  (severity, quiet hours, digest), not in the poller.
+- **Step 3 (embedder relevance) is a seam, not an implementation.** `rss.Ranker`
+  exists and is wired nil. Scoring items against an "interest profile" needs a
+  profile, and there is none yet; a threshold with nothing to compare against is a
+  random filter with a confident name. The filter that runs is the per-feed
+  include/exclude keyword list, which he can read and predict.
+
+No new dependency: stdlib `encoding/xml`, no gofeed. Stock deploy config has no
+`feeds` block, so the capability is off.

docs/plans/14-web-crawler.md

@@ -28,3 +28,41 @@
 7. Add IPC methods `MethodTriggerCrawl(name)`, `MethodListCrawls`, `MethodGetCrawlResult(name)`
 8. Add `crawls` block to `config.Config` and `deploy/mavend.json`
 9. Test with a static HTML page — verify extraction matches expected values, verify scheduling fires correctly
+
+## Shipped 2026-08-01 (#259)
+
+Built as `internal/crawl` (pure: robots, extraction, watcher) plus
+`cmd/mavend/crawls.go` (fetcher, ticker, dedup facts), on top of the guarded
+`internal/webfetch` door added with the feed reader (#258). Off unless
+configured, in two separately-switched halves: `crawl.on_demand` for a URL he
+names, `crawl.watches` for a scheduled re-read.
+
+**Limits are code, not documentation** (`internal/webfetch`, tested one test per
+limit): host allowlist/denylist, no private addresses (loopback, RFC1918 —
+hence the LAN and the `10.42.0.0/24` wg range —, link-local incl. cloud
+metadata, CGNAT, v6 ULA) enforced in the dialer's `Control` hook so DNS
+rebinding and every redirect hop are covered, response size cap, redirect cap,
+timeout, one request per host per second. `robots.txt` is fetched first, cached
+per host, and a `Disallow` is refused with no override.
+
+Deliberate deviations from the plan above:
+
+- **No CSS selectors and no LLM structured extraction** (steps 2). The output is
+  plaintext handed to the phraser as context for the question he asked. A 1.7B
+  extracting a JSON price table from 4000 runes is a worse bet than reading, and
+  `goquery` is not vendored.
+- **No `crawl` act verb and no new IPC methods** (steps 5, 7). Reading a page is
+  a query source (`queryWeb` in `actions_query.go`, last in the chain, behind
+  Kiwix once that is wired), not an action he commands. Nothing needs a new wire
+  method to work.
+- **Notes, not facts.** A page's text is not a fact about him. Only the dedup
+  hash is a fact (`crawl:hash:<name>`, kind `config`, source `poll:crawl`).
+- **Nothing is dispatched.** A changed page writes a note; it does not nudge.
+  Not a nag.
+- **No `/tools` crawl history page.** The notes and the hash facts are already
+  visible on `/dash`.
+
+**No new dependency.** The vendored tree has no `x/net/html`, no `goquery` and
+no `temoto/robotstxt`, so robots parsing and HTML-to-text are stdlib
+(`regexp`, `html`) — RE2 has no backreferences, hence the `pairsRE` builder in
+`extract.go`.

internal/auth/policy.go

@@ -74,7 +74,12 @@ func Requirement(m ipc.Method) Authority {
 		// existing analogue.
 		ipc.MethodCaptureTask,
 		ipc.MethodListTasks,
-		ipc.MethodSetTaskStatus:
+		ipc.MethodSetTaskStatus,
+		// Mail ingestion (Vikunja #246). AuthRead because of what the method can
+		// produce: candidate tasks and nothing else. It cannot write a fact, set a
+		// reminder, or touch the tool allowlist, so a compromised mail reader can
+		// at worst put junk on a review page he clears in one click.
+		ipc.MethodIngestMail:
 		return AuthRead
 	}
 	// Unknown method ⇒ AuthRead, but ipc.dispatch returns ErrUnknownMethod

internal/config/config.go

@@ -150,6 +150,21 @@ type Config struct {
 	// absent ⇒ no evaluation loop at all. See MemoryEvalConfig.
 	MemoryEval *MemoryEvalConfig `json:"memory_eval,omitempty"`
 
+	// Email — mail ingestion (Vikunja #246). nil / absent ⇒ core refuses
+	// ipc.MethodIngestMail outright, so a mail reader cannot make Maven read a
+	// mailbox by merely existing. See EmailConfig; the IMAP host and credential
+	// live in the reader (cmd/mavmaild), never here.
+	Email *EmailConfig `json:"email,omitempty"`
+
+	// Feeds — RSS/Atom feed reading (Vikunja #258). nil / absent ⇒ no feed is
+	// ever fetched: reading the outside world is off unless configured, like
+	// the weather and telegram. See FeedsConfig.
+	Feeds *FeedsConfig `json:"feeds,omitempty"`
+
+	// Crawl — reading a web page (Vikunja #259). nil / absent ⇒ Maven never
+	// fetches a page: not on request, not on a schedule. See CrawlConfig.
+	Crawl *CrawlConfig `json:"crawl,omitempty"`
+
 	// Praxis — the ecosystem attention-state service. When configured, maven
 	// calls the Praxis HTTP tools API for attention listing and item lifecycle.
 	// Maven never touches Praxis's database directly (ecosystem invariant: no
@@ -396,6 +411,109 @@ func (p *PatternProposalConfig) AnnounceProposals() bool {
 	return p != nil && p.Notify
 }
 
+// FeedsConfig — the RSS/Atom reader (Vikunja #258, docs/plans/13-rss-news-feeds.md).
+//
+// Absent ⇒ off, and off means no outbound request at all. Present with an empty
+// `sources` list is also off — a poller with nothing to poll is not wired.
+//
+// What a feed may NOT do here: speak. Items are written as notes with source
+// "rss:<name>" and read back when he asks; nothing is dispatched, nudged or
+// announced on arrival. That is the "not a nag" constraint, and it is why there
+// is no severity or channel field in this block to reach for.
+type FeedsConfig struct {
+	// Sources — the feeds to read. Empty ⇒ the reader stays down.
+	Sources []FeedSourceConfig `json:"sources,omitempty"`
+
+	// PollInterval — default per-feed cadence. 0 ⇒ rss.DefaultPollInterval (30m).
+	PollInterval Duration `json:"poll_interval,omitempty"`
+
+	// MaxItems — most items kept from one feed in one poll. 0 ⇒
+	// rss.DefaultMaxItems (5). This is the "не завали мне /dash" knob.
+	MaxItems int `json:"max_items,omitempty"`
+
+	// MaxAge — on a first poll (no saved mark), how far back to take items.
+	// 0 ⇒ rss.DefaultMaxAge (24h), so switching a feed on imports today, not
+	// the archive.
+	MaxAge Duration `json:"max_age,omitempty"`
+
+	// AllowHosts — when set, the reader may only connect to these hosts (and
+	// their subdomains). The feed URLs' own hosts are added automatically, so
+	// this is only needed to be stricter than that.
+	AllowHosts []string `json:"allow_hosts,omitempty"`
+
+	// Timeout — per-request budget. 0 ⇒ webfetch.DefaultTimeout.
+	Timeout Duration `json:"timeout,omitempty"`
+
+	// MaxBytes — response size cap. 0 ⇒ webfetch.DefaultMaxBytes (2 MiB).
+	MaxBytes int64 `json:"max_bytes,omitempty"`
+}
+
+// FeedSourceConfig — one feed.
+type FeedSourceConfig struct {
+	Name     string   `json:"name"`               // note source is "rss:<name>"
+	URL      string   `json:"url"`                // http(s) only
+	Category string   `json:"category,omitempty"` // "технологии" — what "что нового по X?" matches
+	Interval Duration `json:"interval,omitempty"` // 0 ⇒ FeedsConfig.PollInterval
+	Include  []string `json:"include,omitempty"`  // keep only items containing one of these
+	Exclude  []string `json:"exclude,omitempty"`  // drop items containing any of these
+}
+
+// CrawlConfig — the web crawler (Vikunja #259, docs/plans/14-web-crawler.md).
+//
+// Absent ⇒ off, and off means no page is ever fetched. Present with neither
+// `on_demand` nor a `watches` entry is also off: there would be nothing to do.
+//
+// The crawler is the LAST place an answer is looked for, behind the model, his
+// own memory and the local Kiwix ZIMs. That ordering lives in the query-source
+// chain (cmd/mavend/actions_query.go), not here, but it is the reason this block
+// is small: it is a fallback, not a search engine.
+//
+// Only the URL leaves the box. His notes, facts, persona block and history are
+// never part of a request — the crawler package cannot even read the store.
+type CrawlConfig struct {
+	// OnDemand — may he ask her to read a page he names out loud
+	// ("посмотри https://… — что там пишут?"). false ⇒ the on-demand answer
+	// source stays off and only the watches below run.
+	OnDemand bool `json:"on_demand,omitempty"`
+
+	// Watches — pages re-read on a schedule. A page whose text changed is
+	// written as a note (source "crawl:<name>"); nothing is announced.
+	Watches []CrawlWatchConfig `json:"watches,omitempty"`
+
+	// Interval — default watch cadence. 0 ⇒ crawl.DefaultWatchInterval (6h).
+	Interval Duration `json:"interval,omitempty"`
+
+	// AllowHosts — when set, the ONLY hosts the crawler may reach (subdomains
+	// included). Watched pages' own hosts are added automatically. Setting this
+	// is how "she may read the arch wiki and nothing else" is expressed.
+	AllowHosts []string `json:"allow_hosts,omitempty"`
+
+	// DenyHosts — never reachable, checked first. Private addresses do not need
+	// to be listed: they are refused unconditionally (see internal/webfetch).
+	DenyHosts []string `json:"deny_hosts,omitempty"`
+
+	// UserAgent — sent on every request AND matched against robots.txt groups.
+	// Empty ⇒ webfetch.DefaultUserAgent.
+	UserAgent string `json:"user_agent,omitempty"`
+
+	// Timeout — per-request budget. 0 ⇒ webfetch.DefaultTimeout.
+	Timeout Duration `json:"timeout,omitempty"`
+
+	// MaxBytes — response size cap. 0 ⇒ webfetch.DefaultMaxBytes (2 MiB).
+	MaxBytes int64 `json:"max_bytes,omitempty"`
+
+	// MaxRunes — how much extracted text is kept. 0 ⇒ crawl.DefaultMaxRunes
+	// (4000), which is what fits a 4096-token context alongside a prompt.
+	MaxRunes int `json:"max_runes,omitempty"`
+}
+
+// CrawlWatchConfig — one page kept an eye on.
+type CrawlWatchConfig struct {
+	Name     string   `json:"name"` // note source is "crawl:<name>"
+	URL      string   `json:"url"`
+	Interval Duration `json:"interval,omitempty"` // 0 ⇒ CrawlConfig.Interval
+}
+
 // MemoryEvalConfig — the background memory-evaluation loop (Vikunja #248).
 // Absent ⇒ off, like every other capability that costs something the owner did
 // not ask for. Each evaluation is a full LLM round-trip on the one resident
@@ -418,6 +536,26 @@ type MemoryEvalConfig struct {
 	MinConfidence float64 `json:"min_confidence,omitempty"`
 }
 
+// EmailConfig — core's half of the email reader: how many task candidates one
+// message may produce, and how long the extraction call may take.
+//
+// There is deliberately nothing about a mailbox here. Core does not connect to
+// IMAP, does not know an account exists, and holds no mail credential — the
+// reader daemon does, the same split mavpoll uses for the zenmoney token. This
+// block only says "extraction is allowed, with these bounds".
+type EmailConfig struct {
+	// MaxTasks — candidates per message. 0 ⇒ email.MaxCandidates (3).
+	MaxTasks int `json:"max_tasks,omitempty"`
+
+	// Timeout — per-message extraction budget. 0 ⇒ DefaultEmailTimeout. This is
+	// a Thinking model reading a mail; nobody is waiting on the answer, but a
+	// hung llama-server must not pin the reader's connection forever.
+	Timeout Duration `json:"timeout,omitempty"`
+}
+
+// DefaultEmailTimeout — extraction budget per message.
+const DefaultEmailTimeout = 2 * time.Minute
+
 // PhraserConfig — the LLM-backed phraser seam. The daemon spawns llama-server
 // as a managed subprocess and sends chat-completion requests to phrase nudge
 // and reminder messages. nil ⇒ the template-based Stub is used instead.
@@ -607,6 +745,24 @@ func (c *Config) applyDefaults() {
 		c.MemoryEval.Interval = Duration(DefaultMemoryEvalInterval)
 	}
 
+	// Same rule again: absent stays nil (⇒ mail ingestion refused), present gets
+	// the timeout default so `{}` is a valid "on with the defaults".
+	if c.Email != nil && c.Email.Timeout <= 0 {
+		c.Email.Timeout = Duration(DefaultEmailTimeout)
+	}
+
+	// A feeds block with no sources is the same as no block: nothing to poll,
+	// nothing wired. Normalising it to nil keeps that "off" in one place.
+	if c.Feeds != nil && len(c.Feeds.Sources) == 0 {
+		c.Feeds = nil
+	}
+
+	// Same rule for the crawler: a block that neither answers on demand nor
+	// watches anything has nothing to do, so it is normalised to "off".
+	if c.Crawl != nil && !c.Crawl.OnDemand && len(c.Crawl.Watches) == 0 {
+		c.Crawl = nil
+	}
+
 	if c.Voice != nil {
 		if c.Voice.RouterThreshold <= 0 {
 			c.Voice.RouterThreshold = DefaultRouterThreshold

internal/crawl/crawl.go

@@ -0,0 +1,174 @@
+// Package crawl reads a web page: fetch, robots check, HTML to text.
+//
+// It is the LAST place Maven looks for an answer, and that ordering is the whole
+// design. "Never phones home" is deprecated, but what replaced it puts local
+// sources first: the resident model, then his own memory, then the Kiwix ZIMs on
+// the box (internal/kiwix), and only then the network. A local read costs
+// nothing and leaks nothing; a fetch costs a round-trip and puts a URL in
+// someone's access log. So this package exists to be the fallback, not the
+// front door — see the querySources chain in cmd/mavend/actions_query.go for
+// where it actually sits.
+//
+// What never leaves the box: his notes, his facts, the persona block, the
+// conversation history. Only the URL is requested and, for the on-demand path,
+// only because he said it out loud. Nothing here reads the store.
+//
+// The limits are not in this package — they are in internal/webfetch, which is
+// the only way anything here touches a socket: http(s) only, host allow/deny,
+// private-address refusal, size cap, redirect cap, per-host rate limit. What
+// this package adds is politeness (robots.txt) and dedup.
+package crawl
+
+import (
+	"context"
+	"crypto/sha256"
+	"encoding/hex"
+	"errors"
+	"fmt"
+	"net/url"
+	"strings"
+	"time"
+)
+
+// Errors callers distinguish.
+var (
+	ErrRobots  = errors.New("crawl: robots.txt disallows this path")
+	ErrNotHTML = errors.New("crawl: response is not html or text")
+)
+
+// Fetcher is the guarded HTTP door (internal/webfetch adapted by the daemon). An
+// interface so this package constructs no http.Client of its own and can be
+// tested without a network.
+type Fetcher interface {
+	Get(ctx context.Context, url string) (*Response, error)
+}
+
+// Response is the minimum a crawl needs from a fetch.
+type Response struct {
+	URL         string
+	ContentType string
+	Body        []byte
+}
+
+// Config — crawler knobs.
+type Config struct {
+	// UserAgent is the name matched against robots.txt groups. It must be the
+	// same string the fetcher sends, or Maven would be claiming one identity
+	// and obeying the rules for another.
+	UserAgent string
+	// MaxRunes caps extracted text. 0 ⇒ DefaultMaxRunes.
+	MaxRunes int
+	// RobotsTTL — how long a parsed robots.txt is trusted. 0 ⇒ 1h.
+	RobotsTTL time.Duration
+	// Now is injectable for tests. nil ⇒ time.Now.
+	Now func() time.Time
+}
+
+// Crawler fetches and extracts pages. Safe for concurrent use.
+type Crawler struct {
+	fetch  Fetcher
+	cfg    Config
+	robots *robotsCache
+}
+
+// New builds a crawler. Returns nil when there is no fetcher, which is how the
+// daemon expresses "crawling is off unless configured".
+func New(fetch Fetcher, cfg Config) *Crawler {
+	if fetch == nil {
+		return nil
+	}
+	if cfg.UserAgent == "" {
+		cfg.UserAgent = "Maven"
+	}
+	if cfg.MaxRunes <= 0 {
+		cfg.MaxRunes = DefaultMaxRunes
+	}
+	if cfg.RobotsTTL <= 0 {
+		cfg.RobotsTTL = time.Hour
+	}
+	if cfg.Now == nil {
+		cfg.Now = time.Now
+	}
+	return &Crawler{fetch: fetch, cfg: cfg, robots: newRobotsCache(cfg.RobotsTTL)}
+}
+
+// Page fetches rawURL and returns its text. It checks robots.txt first and
+// refuses a disallowed path with ErrRobots — there is no override.
+func (c *Crawler) Page(ctx context.Context, rawURL string) (Page, error) {
+	u, err := url.Parse(strings.TrimSpace(rawURL))
+	if err != nil {
+		return Page{}, fmt.Errorf("crawl: bad url %q: %w", rawURL, err)
+	}
+	ok, err := c.allowed(ctx, u)
+	if err != nil {
+		return Page{}, err
+	}
+	if !ok {
+		return Page{}, fmt.Errorf("%w: %s", ErrRobots, u.Path)
+	}
+	resp, err := c.fetch.Get(ctx, u.String())
+	if err != nil {
+		return Page{}, err
+	}
+	// A PDF or an image is bytes Maven cannot read; saying so beats storing
+	// binary garbage as a "note".
+	ct := strings.ToLower(resp.ContentType)
+	if ct != "" && !strings.Contains(ct, "html") && !strings.Contains(ct, "text/") &&
+		!strings.Contains(ct, "xml") && !strings.Contains(ct, "json") {
+		return Page{}, fmt.Errorf("%w: %s", ErrNotHTML, resp.ContentType)
+	}
+	return Extract(resp.URL, resp.Body, c.cfg.MaxRunes), nil
+}
+
+// allowed consults robots.txt for u's host, reading it at most once per TTL.
+//
+// A robots.txt that cannot be fetched (404, a timeout, a blocked host) means
+// allow, per the standard. The one thing that is NOT fail-open is an explicit
+// Disallow.
+func (c *Crawler) allowed(ctx context.Context, u *url.URL) (bool, error) {
+	host := u.Host
+	now := c.cfg.Now()
+	rules, ok := c.robots.get(host, now)
+	if !ok {
+		robotsURL := u.Scheme + "://" + host + "/robots.txt"
+		resp, err := c.fetch.Get(ctx, robotsURL)
+		switch {
+		case err != nil:
+			// Note what is NOT swallowed: a refusal from the guarded fetcher.
+			// If webfetch says this host is denied or private, the page fetch
+			// would fail the same way, and reporting the real reason beats
+			// reporting a robots verdict we never got.
+			if isFatalFetchError(err) {
+				return false, err
+			}
+			rules = Rules{}
+		default:
+			rules = ParseRobots(string(resp.Body), c.cfg.UserAgent)
+		}
+		c.robots.put(host, rules, now)
+	}
+	path := u.EscapedPath()
+	if u.RawQuery != "" {
+		path += "?" + u.RawQuery
+	}
+	return rules.Allowed(path), nil
+}
+
+// isFatalFetchError — a fetch failure that means "this host is off limits"
+// rather than "there is no robots.txt here". The sentinel set is webfetch's, but
+// this package must not import it (the interface exists precisely so it does
+// not), so the check is on the message. Ugly and honest: the alternative is a
+// dependency inversion for two strings.
+func isFatalFetchError(err error) bool {
+	s := err.Error()
+	return strings.Contains(s, "not allowed") || strings.Contains(s, "private address") ||
+		strings.Contains(s, "only http and https")
+}
+
+// Hash is the dedup key for a crawl result: the sha256 of the extracted text,
+// hex, first 16 chars. Text and not raw HTML, because a page whose only change
+// is a rotating ad slot or a CSRF token has not changed.
+func Hash(text string) string {
+	sum := sha256.Sum256([]byte(strings.TrimSpace(text)))
+	return hex.EncodeToString(sum[:])[:16]
+}

internal/crawl/crawl_test.go

@@ -0,0 +1,155 @@
+package crawl
+
+import (
+	"context"
+	"errors"
+	"strings"
+	"testing"
+	"time"
+)
+
+// fakeFetcher serves canned pages by URL and counts requests, so a test can
+// assert that robots.txt was read once and that a refusal never reached the page.
+type fakeFetcher struct {
+	pages map[string]Response
+	err   error
+	calls []string
+}
+
+func (f *fakeFetcher) Get(_ context.Context, u string) (*Response, error) {
+	f.calls = append(f.calls, u)
+	if f.err != nil {
+		return nil, f.err
+	}
+	r, ok := f.pages[u]
+	if !ok {
+		return nil, errors.New("http 404")
+	}
+	if r.URL == "" {
+		r.URL = u
+	}
+	if r.ContentType == "" {
+		r.ContentType = "text/html; charset=utf-8"
+	}
+	return &r, nil
+}
+
+const htmlPage = `<html><head><title>Почему небо синее</title>
+<style>body{color:red}</style><script>track()</script></head>
+<body><nav>меню</nav><h1>Небо</h1>
+<p>Свет рассеивается на молекулах воздуха.</p>
+<p>Короткие волны рассеиваются сильнее.</p>
+<footer>© 2026</footer></body></html>`
+
+func newTestCrawler(f *fakeFetcher) *Crawler {
+	return New(f, Config{UserAgent: "Maven/1.0", Now: func() time.Time { return time.Unix(0, 0) }})
+}
+
+func TestPageExtractsText(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{
+		"https://example.org/sky": {Body: []byte(htmlPage)},
+	}}
+	page, err := newTestCrawler(f).Page(context.Background(), "https://example.org/sky")
+	if err != nil {
+		t.Fatal(err)
+	}
+	if page.Title != "Почему небо синее" {
+		t.Errorf("title = %q", page.Title)
+	}
+	if !strings.Contains(page.Text, "Свет рассеивается") {
+		t.Errorf("body text missing: %q", page.Text)
+	}
+	for _, junk := range []string{"track()", "color:red", "меню", "© 2026"} {
+		if strings.Contains(page.Text, junk) {
+			t.Errorf("%q survived extraction: %q", junk, page.Text)
+		}
+	}
+}
+
+func TestRobotsIsCheckedAndObeyed(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{
+		"https://example.org/robots.txt": {Body: []byte("User-agent: *\nDisallow: /secret\n"), ContentType: "text/plain"},
+		"https://example.org/secret/x":   {Body: []byte(htmlPage)},
+		"https://example.org/open":       {Body: []byte(htmlPage)},
+	}}
+	c := newTestCrawler(f)
+	if _, err := c.Page(context.Background(), "https://example.org/secret/x"); !errors.Is(err, ErrRobots) {
+		t.Fatalf("error = %v, want ErrRobots", err)
+	}
+	for _, u := range f.calls {
+		if strings.Contains(u, "/secret") {
+			t.Fatal("the disallowed page was fetched anyway")
+		}
+	}
+	if _, err := c.Page(context.Background(), "https://example.org/open"); err != nil {
+		t.Fatalf("allowed page: %v", err)
+	}
+	// robots.txt was read once for the host, not once per page.
+	robotsReads := 0
+	for _, u := range f.calls {
+		if strings.HasSuffix(u, "/robots.txt") {
+			robotsReads++
+		}
+	}
+	if robotsReads != 1 {
+		t.Fatalf("robots.txt read %d times, want 1", robotsReads)
+	}
+}
+
+// No robots.txt means allow — that is the standard, and the alternative makes
+// most of the web unreadable.
+func TestMissingRobotsAllows(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{
+		"https://example.org/page": {Body: []byte(htmlPage)},
+	}}
+	if _, err := newTestCrawler(f).Page(context.Background(), "https://example.org/page"); err != nil {
+		t.Fatalf("err = %v, want the page", err)
+	}
+}
+
+// A refusal from the guarded fetcher must surface as itself, not be laundered
+// into "no robots.txt, go ahead".
+func TestFetcherRefusalIsNotSwallowed(t *testing.T) {
+	f := &fakeFetcher{err: errors.New("webfetch: refusing to connect to a private address: 127.0.0.1")}
+	_, err := newTestCrawler(f).Page(context.Background(), "http://127.0.0.1:9100/mcp")
+	if err == nil || !strings.Contains(err.Error(), "private address") {
+		t.Fatalf("error = %v, want the fetcher's refusal", err)
+	}
+}
+
+func TestNonTextIsRefused(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{
+		"https://example.org/f.pdf": {Body: []byte("%PDF-1.7"), ContentType: "application/pdf"},
+	}}
+	if _, err := newTestCrawler(f).Page(context.Background(), "https://example.org/f.pdf"); !errors.Is(err, ErrNotHTML) {
+		t.Fatalf("error = %v, want ErrNotHTML", err)
+	}
+}
+
+func TestMaxRunesCapsText(t *testing.T) {
+	long := "<html><body><p>" + strings.Repeat("привет ", 2000) + "</p></body></html>"
+	f := &fakeFetcher{pages: map[string]Response{"https://example.org/l": {Body: []byte(long)}}}
+	c := New(f, Config{MaxRunes: 50})
+	page, err := c.Page(context.Background(), "https://example.org/l")
+	if err != nil {
+		t.Fatal(err)
+	}
+	if n := len([]rune(page.Text)); n > 51 {
+		t.Fatalf("text = %d runes, want the 50-rune cap", n)
+	}
+}
+
+func TestNewWithoutFetcherIsNil(t *testing.T) {
+	if New(nil, Config{}) != nil {
+		t.Fatal("a crawler with no fetcher must be nil — crawling is off unless configured")
+	}
+}
+
+func TestHashIgnoresNothingButText(t *testing.T) {
+	if Hash("a") == Hash("b") {
+		t.Fatal("different text hashed the same")
+	}
+	if Hash(" same \n") != Hash("same") {
+		t.Fatal("surrounding whitespace changed the hash")
+	}
+}

internal/crawl/extract.go

@@ -0,0 +1,106 @@
+package crawl
+
+import (
+	"html"
+	"regexp"
+	"strings"
+)
+
+// HTML → text, with a regexp and no tokenizer.
+//
+// golang.org/x/net/html is not vendored and the network is not assumed, so this
+// is stdlib. That is less of a compromise than it sounds: the unit of context
+// here is a few hundred words for a 4096-token model to read, exactly like the
+// Kiwix snippet, so what matters is dropping script/style/nav noise and keeping
+// paragraph boundaries. A DOM would buy correctness on malformed markup that is
+// then thrown away by truncation anyway.
+//
+// What this deliberately does NOT do: run JavaScript, follow links, or extract
+// structured fields with CSS selectors or an LLM prompt. The plan's step 2 asked
+// for the last of those; see docs/plans/14-web-crawler.md for why it was left
+// out for now.
+
+var (
+	// RE2 has no backreferences, so each tag pair is spelled out rather than
+	// captured and matched against itself.
+	dropRE  = regexp.MustCompile(pairsRE("script", "style", "noscript", "svg", "head", "nav", "footer", "form"))
+	titleRE = regexp.MustCompile(`(?is)<title\b[^>]*>(.*?)</title>`)
+	h1RE    = regexp.MustCompile(`(?is)<h1\b[^>]*>(.*?)</h1>`)
+	// Block-level tags become newlines so paragraphs survive as paragraphs.
+	blockRE   = regexp.MustCompile(`(?is)</?(p|div|br|li|tr|h[1-6]|section|article|blockquote|pre)\b[^>]*>`)
+	tagRE     = regexp.MustCompile(`(?s)<[^>]*>`)
+	commentRE = regexp.MustCompile(`(?s)<!--.*?-->`)
+	spaceRE   = regexp.MustCompile(`[ \t\f\v]+`)
+	blankRE   = regexp.MustCompile(`\n{2,}`)
+)
+
+// pairsRE builds `(?is)<tag …>…</tag>|…` for the given tags.
+func pairsRE(tags ...string) string {
+	parts := make([]string, 0, len(tags))
+	for _, t := range tags {
+		parts = append(parts, `<`+t+`\b[^>]*>.*?</`+t+`>`)
+	}
+	return `(?is)` + strings.Join(parts, "|")
+}
+
+// Page is an extracted page.
+type Page struct {
+	URL   string
+	Title string
+	Text  string // plain text, paragraphs separated by single newlines
+}
+
+// Extract turns a fetched HTML document into a Page. maxRunes caps the text (0 ⇒
+// DefaultMaxRunes); the cap is on runes, not bytes, because a Russian page cut
+// at a byte boundary ends in half a letter.
+func Extract(url string, body []byte, maxRunes int) Page {
+	if maxRunes <= 0 {
+		maxRunes = DefaultMaxRunes
+	}
+	s := string(body)
+	s = commentRE.ReplaceAllString(s, " ")
+
+	title := firstGroup(titleRE, s)
+	if title == "" {
+		title = firstGroup(h1RE, s)
+	}
+
+	s = dropRE.ReplaceAllString(s, "\n")
+	s = blockRE.ReplaceAllString(s, "\n")
+	s = tagRE.ReplaceAllString(s, " ")
+	s = html.UnescapeString(s)
+	s = spaceRE.ReplaceAllString(s, " ")
+
+	var lines []string
+	for _, l := range strings.Split(s, "\n") {
+		if l = strings.TrimSpace(l); l != "" {
+			lines = append(lines, l)
+		}
+	}
+	text := blankRE.ReplaceAllString(strings.Join(lines, "\n"), "\n")
+
+	return Page{URL: url, Title: title, Text: TrimRunes(text, maxRunes)}
+}
+
+// DefaultMaxRunes — how much of a page is kept. ~4000 runes is a long answer's
+// worth of context and still leaves room in a 4096-token window for the prompt
+// and the reply.
+const DefaultMaxRunes = 4000
+
+func firstGroup(re *regexp.Regexp, s string) string {
+	m := re.FindStringSubmatch(s)
+	if len(m) < 2 {
+		return ""
+	}
+	t := tagRE.ReplaceAllString(m[1], " ")
+	return strings.TrimSpace(strings.Join(strings.Fields(html.UnescapeString(t)), " "))
+}
+
+// TrimRunes cuts s to at most max runes, on a rune boundary.
+func TrimRunes(s string, max int) string {
+	r := []rune(s)
+	if len(r) <= max {
+		return s
+	}
+	return strings.TrimSpace(string(r[:max])) + "…"
+}

internal/crawl/robots.go

@@ -0,0 +1,211 @@
+package crawl
+
+import (
+	"regexp"
+	"strings"
+	"sync"
+	"time"
+)
+
+// robots.txt, parsed the small way: no wildcards beyond the two the standard
+// actually defines (`*` inside a path and `$` at the end), no sitemaps, no
+// crawl-delay-per-agent gymnastics. A personal assistant reading a handful of
+// pages does not need a spec-complete implementation; it needs to not be rude,
+// and to be auditable in one sitting.
+//
+// Two rules worth stating because they are choices, not accidents:
+//
+//   - a missing or unreadable robots.txt means ALLOW. That is what the standard
+//     says (404 ⇒ unrestricted), and the alternative would make a site that
+//     simply has no robots.txt unreadable;
+//   - an explicit Disallow means REFUSE, and Maven does not offer an override.
+//     There is no "but he asked me to" flag: the page is not read.
+
+// Rules is a parsed robots.txt for one user-agent.
+type Rules struct {
+	allow    []string
+	disallow []string
+	// Delay is Crawl-delay in seconds when the group named one, 0 otherwise.
+	// The fetcher's own per-host rate limit is the floor; this can only make
+	// Maven slower, never faster.
+	Delay time.Duration
+}
+
+// ParseRobots reads robots.txt and returns the rules that apply to agent.
+//
+// Group selection follows the standard: the most specific matching group wins,
+// which here means an exact user-agent match beats `*`. Lines that are neither
+// are ignored rather than guessed at.
+func ParseRobots(body string, agent string) Rules {
+	agent = strings.ToLower(agent)
+
+	type group struct {
+		agents   []string
+		allow    []string
+		disallow []string
+		delay    time.Duration
+	}
+	var groups []group
+	var cur *group
+	// startNew tracks whether the next User-agent line opens a new group or
+	// joins the current one: consecutive User-agent lines share their rules.
+	startNew := true
+
+	for _, raw := range strings.Split(body, "\n") {
+		line := raw
+		if i := strings.IndexByte(line, '#'); i >= 0 {
+			line = line[:i]
+		}
+		line = strings.TrimSpace(line)
+		if line == "" {
+			continue
+		}
+		key, val, ok := strings.Cut(line, ":")
+		if !ok {
+			continue
+		}
+		key = strings.ToLower(strings.TrimSpace(key))
+		val = strings.TrimSpace(val)
+
+		switch key {
+		case "user-agent":
+			if startNew || cur == nil {
+				groups = append(groups, group{})
+				cur = &groups[len(groups)-1]
+				startNew = false
+			}
+			cur.agents = append(cur.agents, strings.ToLower(val))
+		case "disallow":
+			if cur == nil {
+				continue
+			}
+			startNew = true
+			// "Disallow:" with an empty value allows everything, and is not a
+			// path rule at all.
+			if val != "" {
+				cur.disallow = append(cur.disallow, val)
+			}
+		case "allow":
+			if cur == nil {
+				continue
+			}
+			startNew = true
+			if val != "" {
+				cur.allow = append(cur.allow, val)
+			}
+		case "crawl-delay":
+			if cur == nil {
+				continue
+			}
+			startNew = true
+			if d, err := time.ParseDuration(val + "s"); err == nil && d > 0 {
+				cur.delay = d
+			}
+		}
+	}
+
+	var star, exact *group
+	for i := range groups {
+		for _, a := range groups[i].agents {
+			if a == "*" && star == nil {
+				star = &groups[i]
+			}
+			// A robots.txt names "maven", we send "Maven/1.0 (…)": match on
+			// prefix, which is how every crawler reads this field.
+			if a != "*" && a != "" && strings.HasPrefix(agent, a) {
+				exact = &groups[i]
+			}
+		}
+	}
+	g := exact
+	if g == nil {
+		g = star
+	}
+	if g == nil {
+		return Rules{}
+	}
+	return Rules{allow: g.allow, disallow: g.disallow, Delay: g.delay}
+}
+
+// Allowed reports whether path may be fetched. Longest matching rule wins, and
+// Allow beats Disallow at equal length — the standard's tie-break, and the one
+// that makes "Disallow: /" plus "Allow: /public" mean what it looks like.
+func (r Rules) Allowed(path string) bool {
+	if path == "" {
+		path = "/"
+	}
+	best, allowed := -1, true
+	for _, p := range r.disallow {
+		if n, ok := matchPath(p, path); ok && n > best {
+			best, allowed = n, false
+		}
+	}
+	for _, p := range r.allow {
+		if n, ok := matchPath(p, path); ok && n >= best {
+			best, allowed = n, true
+		}
+	}
+	return allowed
+}
+
+// matchPath applies a robots path pattern and returns the pattern's length as
+// the specificity score. `*` matches any run of characters, `$` anchors the end.
+// A pattern is a PREFIX match otherwise, which is what "Disallow: /admin" means.
+func matchPath(pattern, path string) (int, bool) {
+	score := len(pattern)
+	re, err := robotsRegexp(pattern)
+	if err != nil {
+		return 0, false
+	}
+	return score, re.MatchString(path)
+}
+
+// robotsRegexp turns a robots path pattern into an anchored-at-the-start
+// regexp. Everything but `*` and a trailing `$` is a literal, so the pattern is
+// quoted first and the two metacharacters are put back afterwards.
+func robotsRegexp(pattern string) (*regexp.Regexp, error) {
+	end := ""
+	if strings.HasSuffix(pattern, "$") {
+		pattern = strings.TrimSuffix(pattern, "$")
+		end = "$"
+	}
+	parts := strings.Split(pattern, "*")
+	for i, p := range parts {
+		parts[i] = regexp.QuoteMeta(p)
+	}
+	return regexp.Compile("^" + strings.Join(parts, ".*") + end)
+}
+
+// robotsCache holds parsed rules per host so a crawl of ten pages on one site
+// reads robots.txt once. TTL because a site may change its mind, and a daemon
+// that runs for weeks would otherwise never notice.
+type robotsCache struct {
+	ttl time.Duration
+	mu  sync.Mutex
+	m   map[string]robotsEntry
+}
+
+type robotsEntry struct {
+	rules Rules
+	at    time.Time
+}
+
+func newRobotsCache(ttl time.Duration) *robotsCache {
+	return &robotsCache{ttl: ttl, m: map[string]robotsEntry{}}
+}
+
+func (c *robotsCache) get(host string, now time.Time) (Rules, bool) {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	e, ok := c.m[host]
+	if !ok || now.Sub(e.at) > c.ttl {
+		return Rules{}, false
+	}
+	return e.rules, true
+}
+
+func (c *robotsCache) put(host string, r Rules, now time.Time) {
+	c.mu.Lock()
+	defer c.mu.Unlock()
+	c.m[host] = robotsEntry{rules: r, at: now}
+}

internal/crawl/robots_test.go

@@ -0,0 +1,84 @@
+package crawl
+
+import (
+	"testing"
+	"time"
+)
+
+const robotsBody = `# a comment
+User-agent: *
+Disallow: /private
+Disallow: /tmp/
+Crawl-delay: 5
+
+User-agent: Maven
+Disallow: /
+Allow: /public
+`
+
+func TestParseRobotsPicksTheMostSpecificGroup(t *testing.T) {
+	// The Maven group applies to us even though we send a longer UA string.
+	r := ParseRobots(robotsBody, "Maven/1.0 (self-hosted personal assistant)")
+	if r.Allowed("/anything") {
+		t.Error("Disallow: / in our own group was ignored")
+	}
+	if !r.Allowed("/public/page") {
+		t.Error("Allow: /public must beat the shorter Disallow: /")
+	}
+
+	// A different agent falls into the * group.
+	star := ParseRobots(robotsBody, "SomeoneElse/2")
+	if !star.Allowed("/anything") {
+		t.Error("the * group disallows nothing but /private and /tmp/")
+	}
+	if star.Allowed("/private/x") || star.Allowed("/tmp/") {
+		t.Error("the * group's disallows were not applied")
+	}
+	if star.Delay != 5*time.Second {
+		t.Errorf("crawl-delay = %v, want 5s", star.Delay)
+	}
+}
+
+func TestParseRobotsEmptyMeansAllowAll(t *testing.T) {
+	for _, body := range []string{"", "# nothing here\n", "User-agent: *\nDisallow:\n"} {
+		if !ParseRobots(body, "Maven").Allowed("/whatever") {
+			t.Errorf("body %q must allow everything", body)
+		}
+	}
+}
+
+func TestRobotsWildcards(t *testing.T) {
+	r := ParseRobots("User-agent: *\nDisallow: /*.pdf$\nDisallow: /a/*/secret\n", "Maven")
+	if r.Allowed("/docs/manual.pdf") {
+		t.Error("*.pdf$ did not match")
+	}
+	if !r.Allowed("/docs/manual.pdf.html") {
+		t.Error("$ must anchor at the end")
+	}
+	if r.Allowed("/a/b/secret") {
+		t.Error("/a/*/secret did not match")
+	}
+	if !r.Allowed("/a/b/public") {
+		t.Error("unrelated path was refused")
+	}
+}
+
+// Consecutive User-agent lines share one group, which is common in the wild.
+func TestRobotsSharedGroup(t *testing.T) {
+	r := ParseRobots("User-agent: Googlebot\nUser-agent: Maven\nDisallow: /x\n", "Maven/1.0")
+	if r.Allowed("/x/y") {
+		t.Fatal("a shared group's rules were not applied to the second agent")
+	}
+}
+
+func TestRobotsCacheTTL(t *testing.T) {
+	c := newRobotsCache(time.Minute)
+	now := time.Now()
+	c.put("example.com", ParseRobots("User-agent: *\nDisallow: /\n", "Maven"), now)
+	if _, ok := c.get("example.com", now.Add(30*time.Second)); !ok {
+		t.Error("a fresh entry must be served from cache")
+	}
+	if _, ok := c.get("example.com", now.Add(2*time.Minute)); ok {
+		t.Error("an expired entry must be re-read")
+	}
+}

internal/crawl/watch.go

@@ -0,0 +1,177 @@
+package crawl
+
+import (
+	"context"
+	"fmt"
+	"log"
+	"strings"
+	"time"
+)
+
+// Scheduled crawls: a page is re-read on an interval, and when its TEXT changed
+// the new text is written as a note. Nothing is dispatched — same rule as the
+// feed poller (Vikunja #258). A page that announced its own change would be a
+// nag, and "the docs page changed" is not worth interrupting anyone for.
+//
+// Dedup is by content hash, so a page that re-renders identically writes nothing
+// and a rotating ad slot does not count as news.
+
+// WatchConfig — one page to keep an eye on.
+type WatchConfig struct {
+	Name     string        // note source is "crawl:<Name>"
+	URL      string        // http(s), guarded by the fetcher
+	Interval time.Duration // 0 ⇒ Watcher's default
+}
+
+// Notes is core's note-writing half (same shape as ipc.CoreAPI's method).
+type Notes interface {
+	WriteNote(ctx context.Context, ts time.Time, text string, embedding []float32, source string) (int64, error)
+}
+
+// Hashes remembers the last text hash per watch, durably, so a restart does not
+// re-note an unchanged page. The daemon backs this with config facts
+// ("crawl:hash:<name>").
+type Hashes interface {
+	LastHash(ctx context.Context, name string) (string, error)
+	SetHash(ctx context.Context, name, hash string) error
+}
+
+// Embedder embeds a note on its way into the store. nil ⇒ no vector.
+type Embedder interface {
+	Embed(ctx context.Context, text string) ([]float32, error)
+}
+
+// DefaultWatchInterval — pages change slowly, and every check is a request in
+// someone's log.
+const DefaultWatchInterval = 6 * time.Hour
+
+// Watcher re-reads watched pages on their interval.
+type Watcher struct {
+	c        *Crawler
+	watches  []WatchConfig
+	notes    Notes
+	hashes   Hashes
+	embed    Embedder
+	interval time.Duration
+	nextDue  map[string]time.Time
+}
+
+// NewWatcher wires the scheduled half, or returns nil when there is nothing to
+// watch. Callers check for nil: no watches, no goroutine, no request.
+func NewWatcher(c *Crawler, watches []WatchConfig, notes Notes, hashes Hashes, embed Embedder, defaultInterval time.Duration) *Watcher {
+	if c == nil || notes == nil {
+		return nil
+	}
+	var valid []WatchConfig
+	for _, w := range watches {
+		if strings.TrimSpace(w.Name) == "" || strings.TrimSpace(w.URL) == "" {
+			log.Printf("crawl: skipping a watch with no name or no url")
+			continue
+		}
+		valid = append(valid, w)
+	}
+	if len(valid) == 0 {
+		return nil
+	}
+	if defaultInterval <= 0 {
+		defaultInterval = DefaultWatchInterval
+	}
+	return &Watcher{
+		c: c, watches: valid, notes: notes, hashes: hashes, embed: embed,
+		interval: defaultInterval, nextDue: map[string]time.Time{},
+	}
+}
+
+// Watches returns the configured watches.
+func (w *Watcher) Watches() []WatchConfig { return w.watches }
+
+// CheckDue re-reads every watch whose interval elapsed and returns how many
+// notes were written. Errors are logged per watch, never returned: one dead page
+// must not stop the others.
+func (w *Watcher) CheckDue(ctx context.Context, now time.Time) int {
+	written := 0
+	for _, watch := range w.watches {
+		if due, ok := w.nextDue[watch.Name]; ok && now.Before(due) {
+			continue
+		}
+		interval := watch.Interval
+		if interval <= 0 {
+			interval = w.interval
+		}
+		w.nextDue[watch.Name] = now.Add(interval)
+		changed, err := w.Check(ctx, watch, now)
+		if err != nil {
+			log.Printf("crawl: watch %s: %v", watch.Name, err)
+			continue
+		}
+		if changed {
+			log.Printf("crawl: watch %s: page changed, noted", watch.Name)
+			written++
+		}
+	}
+	return written
+}
+
+// Check re-reads one watch now and reports whether it wrote a note.
+func (w *Watcher) Check(ctx context.Context, watch WatchConfig, now time.Time) (bool, error) {
+	page, err := w.c.Page(ctx, watch.URL)
+	if err != nil {
+		return false, err
+	}
+	// Title included: a page whose headline changed has changed.
+	h := Hash(page.Title + "\n" + page.Text)
+	if w.hashes != nil {
+		prev, err := w.hashes.LastHash(ctx, watch.Name)
+		if err != nil {
+			log.Printf("crawl: watch %s: read hash: %v", watch.Name, err)
+		}
+		if prev == h {
+			return false, nil
+		}
+	}
+	text := NoteText(watch, page)
+	var vec []float32
+	if w.embed != nil {
+		v, err := w.embed.Embed(ctx, text)
+		if err != nil {
+			log.Printf("crawl: watch %s: embed: %v", watch.Name, err)
+		} else {
+			vec = v
+		}
+	}
+	if _, err := w.notes.WriteNote(ctx, now, text, vec, SourceFor(watch.Name)); err != nil {
+		return false, fmt.Errorf("write note: %w", err)
+	}
+	if w.hashes != nil {
+		if err := w.hashes.SetHash(ctx, watch.Name, h); err != nil {
+			log.Printf("crawl: watch %s: save hash: %v", watch.Name, err)
+		}
+	}
+	return true, nil
+}
+
+// SourceFor is the note source for a watch, and SourcePrefix is what the answer
+// path matches to recognise one.
+func SourceFor(name string) string { return SourcePrefix + name }
+
+// SourcePrefix — provenance for anything read off the network on a schedule.
+const SourcePrefix = "crawl:"
+
+// noteRunes — how much of a watched page goes into a note. Shorter than what the
+// on-demand path reads: a note is a record of a change, not an archive.
+const noteRunes = 800
+
+// NoteText renders a watched page as a note body.
+func NoteText(watch WatchConfig, page Page) string {
+	var b strings.Builder
+	if page.Title != "" {
+		b.WriteString(page.Title)
+	} else {
+		b.WriteString(watch.Name)
+	}
+	b.WriteString("\n")
+	b.WriteString(TrimRunes(page.Text, noteRunes))
+	b.WriteString("\n")
+	b.WriteString(watch.URL)
+	return b.String()
+}

internal/crawl/watch_test.go

@@ -0,0 +1,117 @@
+package crawl
+
+import (
+	"context"
+	"strings"
+	"testing"
+	"time"
+)
+
+type note struct {
+	text   string
+	source string
+}
+
+type fakeNotes struct{ notes []note }
+
+func (n *fakeNotes) WriteNote(_ context.Context, _ time.Time, text string, _ []float32, source string) (int64, error) {
+	n.notes = append(n.notes, note{text, source})
+	return int64(len(n.notes)), nil
+}
+
+type fakeHashes struct{ m map[string]string }
+
+func newHashes() *fakeHashes { return &fakeHashes{m: map[string]string{}} }
+func (f *fakeHashes) LastHash(_ context.Context, name string) (string, error) {
+	return f.m[name], nil
+}
+func (f *fakeHashes) SetHash(_ context.Context, name, h string) error { f.m[name] = h; return nil }
+
+var t0 = time.Date(2026, 8, 1, 9, 0, 0, 0, time.UTC)
+
+func TestWatchNotesAChangedPage(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{
+		"https://example.org/docs": {Body: []byte(htmlPage)},
+	}}
+	notes := &fakeNotes{}
+	hashes := newHashes()
+	w := NewWatcher(newTestCrawler(f), []WatchConfig{{Name: "docs", URL: "https://example.org/docs"}},
+		notes, hashes, nil, time.Hour)
+	if w == nil {
+		t.Fatal("NewWatcher returned nil for a configured watch")
+	}
+	if n := w.CheckDue(context.Background(), t0); n != 1 {
+		t.Fatalf("first check wrote %d notes, want 1", n)
+	}
+	if notes.notes[0].source != "crawl:docs" {
+		t.Errorf("source = %q, want crawl:docs", notes.notes[0].source)
+	}
+	if !strings.Contains(notes.notes[0].text, "https://example.org/docs") {
+		t.Errorf("note does not carry the url: %q", notes.notes[0].text)
+	}
+
+	// Unchanged page, interval elapsed: nothing written.
+	if n := w.CheckDue(context.Background(), t0.Add(2*time.Hour)); n != 0 {
+		t.Fatalf("an unchanged page wrote %d notes", n)
+	}
+
+	// Changed page: one note.
+	f.pages["https://example.org/docs"] = Response{Body: []byte(strings.Replace(htmlPage, "синее", "серое", 1))}
+	if n := w.CheckDue(context.Background(), t0.Add(4*time.Hour)); n != 1 {
+		t.Fatalf("a changed page wrote %d notes, want 1", n)
+	}
+}
+
+func TestWatchIntervalIsRespected(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{"https://example.org/d": {Body: []byte(htmlPage)}}}
+	w := NewWatcher(newTestCrawler(f), []WatchConfig{{Name: "d", URL: "https://example.org/d", Interval: time.Hour}},
+		&fakeNotes{}, newHashes(), nil, 0)
+	w.CheckDue(context.Background(), t0)
+	before := len(f.calls)
+	w.CheckDue(context.Background(), t0.Add(time.Minute))
+	if len(f.calls) != before {
+		t.Fatal("the page was re-read inside its interval")
+	}
+}
+
+// The hash is durable so a restart does not re-note an unchanged page.
+func TestWatchHashSurvivesRestart(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{"https://example.org/d": {Body: []byte(htmlPage)}}}
+	hashes := newHashes()
+	watches := []WatchConfig{{Name: "d", URL: "https://example.org/d"}}
+	NewWatcher(newTestCrawler(f), watches, &fakeNotes{}, hashes, nil, time.Hour).CheckDue(context.Background(), t0)
+
+	notes2 := &fakeNotes{}
+	NewWatcher(newTestCrawler(f), watches, notes2, hashes, nil, time.Hour).CheckDue(context.Background(), t0.Add(time.Hour))
+	if len(notes2.notes) != 0 {
+		t.Fatalf("a fresh watcher re-noted an unchanged page: %q", notes2.notes[0].text)
+	}
+}
+
+func TestWatchDeadPageDoesNotStopTheOthers(t *testing.T) {
+	f := &fakeFetcher{pages: map[string]Response{"https://example.org/live": {Body: []byte(htmlPage)}}}
+	notes := &fakeNotes{}
+	w := NewWatcher(newTestCrawler(f), []WatchConfig{
+		{Name: "dead", URL: "https://example.org/gone"},
+		{Name: "live", URL: "https://example.org/live"},
+	}, notes, newHashes(), nil, time.Hour)
+	if n := w.CheckDue(context.Background(), t0); n != 1 {
+		t.Fatalf("wrote %d notes, want 1 (the live page)", n)
+	}
+	if notes.notes[0].source != "crawl:live" {
+		t.Fatalf("source = %q", notes.notes[0].source)
+	}
+}
+
+func TestNoWatchesMeansNoWatcher(t *testing.T) {
+	c := newTestCrawler(&fakeFetcher{})
+	if NewWatcher(c, nil, &fakeNotes{}, nil, nil, 0) != nil {
+		t.Fatal("no watches must mean no watcher")
+	}
+	if NewWatcher(nil, []WatchConfig{{Name: "a", URL: "u"}}, &fakeNotes{}, nil, nil, 0) != nil {
+		t.Fatal("no crawler must mean no watcher")
+	}
+	if NewWatcher(c, []WatchConfig{{Name: "", URL: ""}}, &fakeNotes{}, nil, nil, 0) != nil {
+		t.Fatal("a watch with no name or url is not a configuration")
+	}
+}

internal/email/extract.go

@@ -0,0 +1,226 @@
+package email
+
+import (
+	"context"
+	"encoding/json"
+	"fmt"
+	"strings"
+	"time"
+
+	"github.com/kami/maven/internal/llm"
+	"github.com/kami/maven/internal/persona"
+)
+
+// Extraction — turning one mail into task CANDIDATES, and nothing else.
+//
+// The output of this file can only ever become rows in `tasks` with status
+// "candidate" (store.TaskCandidate), written through the one intake seam
+// (ipc.CaptureTaskReq, Vikunja #130). That bound is the whole design:
+//
+//   - No reminder. A reminder FIRES; it speaks to him unprompted. A 1.7B that
+//     misreads "встреча была в четверг" as a future appointment would then wake
+//     him up about it. A candidate that is wrong is a line on a review page he
+//     dismisses in one click, which is the correct cost of a model being wrong
+//     about someone's mail.
+//   - No fact. A fact is a claim Maven will later recite as true. Nothing read
+//     out of a marketing mail deserves that standing.
+//   - No calendar event, no note, no action. Extraction writes candidates or
+//     writes nothing.
+//
+// The due date the model may return is stored on the candidate (tasks.due_ts),
+// which no scheduler reads — it is there so the review page can sort by it.
+//
+// Privacy: the mail text goes to the resident model on this box and nowhere
+// else. It is never search input (CLAUDE.md: "his notes and facts are never
+// search input" — mail is the same class), and Evidence keeps only the subject
+// line, so the review page shows him where a candidate came from without the
+// store growing a copy of his mailbox.
+
+// MaxCandidates — at most this many candidates per message, enforced by the
+// grammar. A mail with four tasks in it is a mail he has to read himself; a
+// model allowed ten will produce ten.
+const MaxCandidates = 3
+
+// SourcePrefix — provenance for everything this package captures. The mailbox
+// name is appended: "email:INBOX". Same vocabulary as tap:voice / poll:netdata.
+const SourcePrefix = "email:"
+
+// Candidate — one piece of work the model thinks the mail is asking for.
+type Candidate struct {
+	Text string `json:"text"`
+	// Due — "YYYY-MM-DD" or empty. A date the model read out of the text, not a
+	// date it computed: relative wording ("до пятницы") is left in Text, because
+	// a small model resolving "пятница" against today's date gets it wrong often
+	// enough that a stored wrong date is worse than no date.
+	Due string `json:"due"`
+}
+
+// Completer — the llama-server seam, same shape memeval and the router use, so
+// the one resident model serves this caller too.
+type Completer interface {
+	Complete(ctx context.Context, r llm.Req) (string, error)
+}
+
+// Extractor reads a message and returns candidates. It holds no store and no
+// writer on purpose: this type cannot persist anything, so "extraction never
+// acts" is a property of the code, not of a review.
+type Extractor struct {
+	llm Completer
+	// MaxCandidates — 0 ⇒ MaxCandidates.
+	max int
+	// ContextBlock — the shared persona block, optional. Extraction output is
+	// not spoken, so the persona matters less here than in the phraser; it is
+	// wired anyway so a candidate reads in her voice on the review page.
+	contextBlock func() string
+}
+
+func NewExtractor(c Completer, max int, contextBlock func() string) *Extractor {
+	if max <= 0 || max > MaxCandidates {
+		max = MaxCandidates
+	}
+	return &Extractor{llm: c, max: max, contextBlock: contextBlock}
+}
+
+// extractGrammar — GBNF pinning the answer to a bounded array of fixed-shape
+// candidates. Same reasoning as memeval's evalGrammar and the router's
+// routeGrammar: the shape and the length bound are what keep a small model from
+// drifting into prose or spending the token budget repeating one field.
+//
+// The empty array is reachable, deliberately: most mail contains no task, and a
+// model with no way to say "nothing" invents something.
+const extractGrammar = `
+root  ::= "[" ws (item ("," ws item){0,2})? ws "]"
+item  ::= "{" ws "\"text\"" ws ":" ws text "," ws "\"due\"" ws ":" ws due ws "}"
+text  ::= "\"" ([^"\\] | "\\" .){1,120} "\""
+due   ::= "\"\"" | "\"" [0-9]{4} "-" [0-9]{2} "-" [0-9]{2} "\""
+ws    ::= [ \t\n]*
+`
+
+// extractSystem — the extraction prompt.
+//
+// Written around the two failure modes a small model has on this task: it
+// summarises when asked to extract (turning a mail into "письмо от Антона"),
+// and it invents an obligation from any polite closing sentence. Hence the
+// insistence on a verb phrase, and the explicit permission to return [].
+const extractSystem = `Ты читаешь одно письмо из его почты и достаёшь из него дела, которые письмо от него требует.
+
+Правила:
+- Отвечай ТОЛЬКО массивом JSON. Каждый элемент: {"text": "...", "due": "ГГГГ-ММ-ДД" или ""}.
+- text — короткая формулировка дела по-русски, с глаголом: "оплатить счёт за интернет", "отправить акт". Не пересказывай письмо и не описывай его.
+- Дело — это то, что должен сделать ОН. Рассылка, реклама, уведомление, отчёт, письмо «просто к сведению» — дел не содержат.
+- Если письмо ничего от него не требует, верни пустой массив []. Это нормальный ответ, так бывает чаще всего.
+- Ничего не придумывай. Если срока в письме нет — "".
+- due заполняй только когда в письме стоит конкретная дата. Слова вроде «до пятницы» оставь в text, дату не вычисляй.
+- Максимум три дела. Лучше одно точное, чем три общих.`
+
+// Extract returns the candidates in one message.
+//
+// Junk is refused without an LLM call — cheapest possible defence, and the
+// reason the header filter exists. An empty message (no subject, no body) is
+// likewise not worth a round trip.
+//
+// A parse failure is an error the caller logs and moves past. It is never
+// silently turned into zero candidates, because "the model went off the rails"
+// and "the mail contains no task" want different reactions from a human reading
+// the log.
+func (e *Extractor) Extract(ctx context.Context, msg Message) ([]Candidate, error) {
+	if msg.Junk {
+		return nil, nil
+	}
+	user := renderForModel(msg)
+	if user == "" {
+		return nil, nil
+	}
+	raw, err := e.llm.Complete(ctx, llm.Req{
+		System:        persona.Prepend(e.contextBlock, extractSystem),
+		User:          user,
+		Grammar:       extractGrammar,
+		MaxTokens:     512,
+		RepeatPenalty: 1.1,
+	})
+	if err != nil {
+		return nil, fmt.Errorf("email: extract: %w", err)
+	}
+	items, err := parseCandidates(raw)
+	if err != nil {
+		// The raw reply is NOT in the error: it is a transformation of his mail,
+		// and this error reaches the daemon log.
+		return nil, fmt.Errorf("email: extract: unparsable reply (%d bytes)", len(raw))
+	}
+	out := make([]Candidate, 0, len(items))
+	seen := map[string]bool{}
+	for _, it := range items {
+		it.Text = strings.TrimSpace(it.Text)
+		if it.Text == "" {
+			continue
+		}
+		key := strings.ToLower(strings.Join(strings.Fields(it.Text), " "))
+		if seen[key] {
+			continue // the model repeating itself is not two tasks
+		}
+		seen[key] = true
+		if _, ok := ParseDue(it.Due); !ok {
+			it.Due = "" // a date the grammar allowed but the calendar does not
+		}
+		out = append(out, it)
+		if len(out) >= e.max {
+			break
+		}
+	}
+	return out, nil
+}
+
+// renderForModel is the user turn: subject, sender and body, labelled. Only
+// these three fields — no headers, no recipient list, no message-id, nothing
+// that would let the model start reasoning about routing metadata.
+func renderForModel(msg Message) string {
+	var b strings.Builder
+	if msg.From != "" {
+		fmt.Fprintf(&b, "От: %s\n", msg.From)
+	}
+	if msg.Subject != "" {
+		fmt.Fprintf(&b, "Тема: %s\n", msg.Subject)
+	}
+	if msg.Body != "" {
+		fmt.Fprintf(&b, "\n%s\n", msg.Body)
+	}
+	if msg.Subject == "" && msg.Body == "" {
+		return ""
+	}
+	return b.String()
+}
+
+// parseCandidates decodes the grammar-constrained reply, tolerating the
+// wrappers a Thinking model sometimes leaves around it (a fenced block, or
+// leading reasoning before the array).
+func parseCandidates(raw string) ([]Candidate, error) {
+	s := strings.TrimSpace(raw)
+	if i := strings.Index(s, "["); i > 0 {
+		s = s[i:]
+	}
+	if j := strings.LastIndex(s, "]"); j >= 0 {
+		s = s[:j+1]
+	}
+	var out []Candidate
+	if err := json.Unmarshal([]byte(s), &out); err != nil {
+		return nil, err
+	}
+	return out, nil
+}
+
+// ParseDue turns the model's "YYYY-MM-DD" into a time in UTC. Exported because
+// the daemon-side intake stores it on the candidate.
+//
+// The zero-value/empty case returns ok=false rather than an error: no date is
+// the common answer, not a failure.
+func ParseDue(s string) (time.Time, bool) {
+	s = strings.TrimSpace(s)
+	if s == "" {
+		return time.Time{}, false
+	}
+	t, err := time.Parse("2006-01-02", s)
+	if err != nil {
+		return time.Time{}, false
+	}
+	return t, true
+}

internal/email/extract_test.go

@@ -0,0 +1,143 @@
+package email
+
+import (
+	"context"
+	"strings"
+	"testing"
+
+	"github.com/kami/maven/internal/llm"
+)
+
+// fakeLLM returns a canned reply and records the request, so a test can assert
+// on the grammar and on what of the mail was sent.
+type fakeLLM struct {
+	reply string
+	err   error
+	got   llm.Req
+	calls int
+}
+
+func (f *fakeLLM) Complete(_ context.Context, r llm.Req) (string, error) {
+	f.calls++
+	f.got = r
+	return f.reply, f.err
+}
+
+func msgFor(subject, body string) Message {
+	return Message{UID: 1, From: "anton@example.org", Subject: subject, Body: body}
+}
+
+func TestExtractCandidates(t *testing.T) {
+	f := &fakeLLM{reply: `[{"text":"отправить акт","due":""},{"text":"оплатить счёт","due":"2026-08-05"}]`}
+	e := NewExtractor(f, 0, nil)
+	got, err := e.Extract(context.Background(), msgFor("Акт и счёт", "Надо отправить акт и оплатить счёт до 5 августа."))
+	if err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if len(got) != 2 {
+		t.Fatalf("got %d candidates, want 2: %+v", len(got), got)
+	}
+	if got[0].Text != "отправить акт" || got[1].Due != "2026-08-05" {
+		t.Errorf("candidates = %+v", got)
+	}
+	if f.got.Grammar == "" {
+		t.Error("extraction must be grammar-constrained")
+	}
+	// The subject and body go to the model; nothing else about the message does.
+	if !strings.Contains(f.got.User, "Акт и счёт") || !strings.Contains(f.got.User, "оплатить счёт") {
+		t.Errorf("user turn = %q", f.got.User)
+	}
+}
+
+func TestExtractEmptyArrayIsNotAnError(t *testing.T) {
+	f := &fakeLLM{reply: "[]"}
+	got, err := NewExtractor(f, 0, nil).Extract(context.Background(), msgFor("FYI", "Просто к сведению."))
+	if err != nil || len(got) != 0 {
+		t.Fatalf("got (%v, %v), want (empty, nil) — no task is the normal answer", got, err)
+	}
+}
+
+// Junk must never reach the model: the header filter exists so the resident
+// model is not spent on newsletters.
+func TestExtractSkipsJunkWithoutCallingModel(t *testing.T) {
+	f := &fakeLLM{reply: `[{"text":"купить всё со скидкой","due":""}]`}
+	msg := msgFor("Скидки", "Sale!")
+	msg.Junk = true
+	got, err := NewExtractor(f, 0, nil).Extract(context.Background(), msg)
+	if err != nil || got != nil {
+		t.Fatalf("got (%v, %v), want (nil, nil)", got, err)
+	}
+	if f.calls != 0 {
+		t.Errorf("model called %d times for junk, want 0", f.calls)
+	}
+}
+
+func TestExtractEmptyMessageIsNotSent(t *testing.T) {
+	f := &fakeLLM{reply: "[]"}
+	if _, err := NewExtractor(f, 0, nil).Extract(context.Background(), Message{UID: 3}); err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if f.calls != 0 {
+		t.Errorf("model called %d times for an empty message, want 0", f.calls)
+	}
+}
+
+func TestExtractCaps(t *testing.T) {
+	f := &fakeLLM{reply: `[{"text":"a","due":""},{"text":"b","due":""},{"text":"c","due":""}]`}
+	got, err := NewExtractor(f, 2, nil).Extract(context.Background(), msgFor("s", "b"))
+	if err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if len(got) != 2 {
+		t.Errorf("got %d, want the configured cap of 2", len(got))
+	}
+}
+
+func TestExtractDropsRepeatsAndBadDates(t *testing.T) {
+	f := &fakeLLM{reply: `[{"text":"Отправить акт","due":"2026-02-31"},{"text":"отправить  акт","due":""},{"text":"  ","due":""}]`}
+	got, err := NewExtractor(f, 0, nil).Extract(context.Background(), msgFor("s", "b"))
+	if err != nil {
+		t.Fatalf("extract: %v", err)
+	}
+	if len(got) != 1 {
+		t.Fatalf("got %d candidates, want 1 (repeat and blank dropped): %+v", len(got), got)
+	}
+	if got[0].Due != "" {
+		t.Errorf("due = %q, want empty — 2026-02-31 is not a date", got[0].Due)
+	}
+}
+
+// A Thinking model sometimes wraps the array; and when it emits something
+// unparsable the caller must hear about it rather than see "no tasks".
+func TestParseCandidatesTolerance(t *testing.T) {
+	got, err := parseCandidates("думаю... [{\"text\":\"x\",\"due\":\"\"}] всё")
+	if err != nil || len(got) != 1 || got[0].Text != "x" {
+		t.Fatalf("got (%+v, %v)", got, err)
+	}
+	if _, err := parseCandidates("нет никакого JSON"); err == nil {
+		t.Error("unparsable output must be an error")
+	}
+}
+
+func TestExtractParseErrorHidesMailText(t *testing.T) {
+	f := &fakeLLM{reply: "он просил отправить акт, вот такой ответ"}
+	_, err := NewExtractor(f, 0, nil).Extract(context.Background(), msgFor("Акт", "секретный текст"))
+	if err == nil {
+		t.Fatal("want an error")
+	}
+	if strings.Contains(err.Error(), "акт") || strings.Contains(err.Error(), "секретный") {
+		t.Errorf("error text leaks mail content: %v", err)
+	}
+}
+
+func TestParseDue(t *testing.T) {
+	if _, ok := ParseDue(""); ok {
+		t.Error("empty due must be (zero, false)")
+	}
+	if got, ok := ParseDue("2026-08-05"); !ok || got.Year() != 2026 || got.Month() != 8 || got.Day() != 5 {
+		t.Errorf("ParseDue = (%v, %v)", got, ok)
+	}
+	if _, ok := ParseDue("05.08.2026"); ok {
+		t.Error("a non-ISO date must not parse")
+	}
+}

internal/email/fetch.go

@@ -0,0 +1,98 @@
+package email
+
+import (
+	"fmt"
+	"time"
+)
+
+// FetchSince is the whole read path in one call: connect, log in, examine the
+// mailbox read-only, list what arrived since a date, fetch and parse the ones
+// the caller has not seen, log out.
+//
+// It is a function rather than a long-lived object because a mail poller should
+// not hold an authenticated session (and therefore his credential in a live TLS
+// state) between polls. Connect, read, drop.
+//
+// skip decides which UIDs are already known — the poller's seen-set. max bounds
+// one poll: a mailbox that received 400 messages overnight must not turn into
+// 400 LLM calls, and the newest max are the ones a task could still be hiding
+// in. Junk messages are returned too, flagged, so the caller can mark them seen
+// without a second protocol round.
+type FetchSince struct {
+	Addr    string // host or host:993
+	User    string
+	Mailbox string // e.g. "INBOX"
+	Timeout time.Duration
+	Since   time.Time
+	Max     int
+	Skip    func(uid uint32) bool
+}
+
+// Run performs one read. password is passed here, not stored in the struct, so
+// the configuration of a mailbox and the secret for it are never the same value
+// sitting in the same place.
+func (f FetchSince) Run(password string) ([]Message, error) {
+	return f.RunWith(password, nil)
+}
+
+// RunWith is Run with an explicit connection function, which is how the reader
+// daemon and the tests substitute an in-process server. nil ⇒ Dial, i.e.
+// implicit TLS with certificate verification; there is no configuration path
+// that reaches this, so no deployment can end up talking cleartext IMAP.
+func (f FetchSince) RunWith(password string, dial func(addr string, timeout time.Duration) (*Conn, error)) ([]Message, error) {
+	if f.Addr == "" || f.User == "" || f.Mailbox == "" {
+		return nil, fmt.Errorf("email: mailbox not configured (addr/user/mailbox)")
+	}
+	if dial == nil {
+		dial = Dial
+	}
+	c, err := dial(f.Addr, f.Timeout)
+	if err != nil {
+		return nil, err
+	}
+	defer c.Close()
+	if err := c.Login(f.User, password); err != nil {
+		return nil, err
+	}
+	defer c.Logout()
+	if err := c.Select(f.Mailbox); err != nil {
+		return nil, err
+	}
+	uids, err := c.SearchSince(f.Since)
+	if err != nil {
+		return nil, err
+	}
+
+	// Newest UIDs first — IMAP hands them back ascending, and when Max clips the
+	// list the recent mail is what matters.
+	wanted := make([]uint32, 0, len(uids))
+	for i := len(uids) - 1; i >= 0; i-- {
+		if f.Skip != nil && f.Skip(uids[i]) {
+			continue
+		}
+		wanted = append(wanted, uids[i])
+		if f.Max > 0 && len(wanted) >= f.Max {
+			break
+		}
+	}
+
+	out := make([]Message, 0, len(wanted))
+	for _, uid := range wanted {
+		raw, err := c.Fetch(uid)
+		if err != nil {
+			// One unreadable message does not abandon the poll; the rest of the
+			// mailbox is still worth reading. The error names the UID, not the
+			// message.
+			return out, fmt.Errorf("email: fetch uid %d: %w", uid, err)
+		}
+		if len(raw) == 0 {
+			continue // vanished between SEARCH and FETCH
+		}
+		msg, err := ParseMessage(uid, raw)
+		if err != nil {
+			continue // unparsable headers — nothing to review, skip silently
+		}
+		out = append(out, msg)
+	}
+	return out, nil
+}

internal/email/fetch_test.go

@@ -0,0 +1,49 @@
+package email
+
+import (
+	"net"
+	"strings"
+	"testing"
+	"time"
+)
+
+func TestFetchSinceRun(t *testing.T) {
+	mk := func(subject string) string {
+		return "Subject: " + subject + "\r\nContent-Type: text/plain; charset=utf-8\r\n\r\nbody\r\n"
+	}
+	f := &fakeIMAP{
+		uids: []uint32{1, 2, 3},
+		msgs: map[uint32]string{1: mk("one"), 2: mk("two"), 3: mk("three")},
+	}
+	fs := FetchSince{
+		Addr: "mail.example:993", User: "kami", Mailbox: "INBOX",
+		Timeout: 5 * time.Second,
+		Since:   time.Date(2026, 7, 30, 0, 0, 0, 0, time.UTC),
+		Max:     2,
+		Skip:    func(uid uint32) bool { return uid == 3 },
+	}
+	msgs, err := fs.RunWith("secret", func(addr string, timeout time.Duration) (*Conn, error) {
+		cli, srv := net.Pipe()
+		go f.serve(t, srv)
+		return NewConn(cli, timeout)
+	})
+	if err != nil {
+		t.Fatalf("run: %v", err)
+	}
+	// Newest first, the already-seen UID skipped, Max respected.
+	if len(msgs) != 2 {
+		t.Fatalf("got %d messages, want 2: %+v", len(msgs), msgs)
+	}
+	if msgs[0].Subject != "two" || msgs[1].Subject != "one" {
+		t.Errorf("subjects = %q,%q, want two,one (newest first)", msgs[0].Subject, msgs[1].Subject)
+	}
+	if strings.Contains(strings.Join(f.cmds, " "), "UID FETCH 3") {
+		t.Error("a skipped UID must not be fetched again")
+	}
+}
+
+func TestFetchSinceRequiresConfig(t *testing.T) {
+	if _, err := (FetchSince{}).Run("secret"); err == nil {
+		t.Fatal("an unconfigured mailbox must not be read")
+	}
+}

internal/email/imap.go

@@ -0,0 +1,280 @@
+package email
+
+import (
+	"bufio"
+	"crypto/tls"
+	"fmt"
+	"io"
+	"net"
+	"regexp"
+	"strconv"
+	"strings"
+	"time"
+)
+
+// A minimal IMAP4rev1 client — LOGIN, SELECT, UID SEARCH, UID FETCH with
+// BODY.PEEK, LOGOUT, and nothing else.
+//
+// Why hand-rolled instead of go-imap: the whole surface Maven needs is five
+// commands, and this is the one code path that holds his mailbox credential and
+// reads his private mail. A ~200-line client with no dependencies is auditable
+// in one sitting; a general-purpose IMAP library is a much larger amount of
+// code doing much more than we asked, in the most sensitive place in the tree.
+// If IDLE, CONDSTORE or server-side threading ever become worth having, that
+// trade should be re-made deliberately.
+//
+// BODY.PEEK[] rather than BODY[] is load-bearing: Maven reads his mail and must
+// leave no trace of having done so. Reading a message here does not mark it
+// \Seen, so the unread state in his own mail client stays his.
+
+// DefaultIMAPPort — implicit-TLS IMAP. There is no cleartext and no STARTTLS
+// path in this client: an option to send his password over a plain socket is an
+// option to get it wrong once.
+const DefaultIMAPPort = "993"
+
+// Conn — one authenticated IMAP connection. Not safe for concurrent use; the
+// poller drives one connection at a time.
+type Conn struct {
+	rwc     io.ReadWriteCloser
+	r       *bufio.Reader
+	tag     int
+	timeout time.Duration
+}
+
+// Dial opens an implicit-TLS connection and reads the server greeting.
+func Dial(addr string, timeout time.Duration) (*Conn, error) {
+	host, _, err := net.SplitHostPort(addr)
+	if err != nil {
+		host, addr = addr, net.JoinHostPort(addr, DefaultIMAPPort)
+	}
+	d := &net.Dialer{Timeout: timeout}
+	// ServerName is set from the host we asked for: certificate verification is
+	// the only thing standing between his password and a MITM on the way out.
+	c, err := tls.DialWithDialer(d, "tcp", addr, &tls.Config{ServerName: host, MinVersion: tls.VersionTLS12})
+	if err != nil {
+		return nil, fmt.Errorf("email: dial %s: %w", addr, err)
+	}
+	return NewConn(c, timeout)
+}
+
+// NewConn wraps an already-open stream (the tests speak IMAP over a pipe) and
+// consumes the greeting.
+func NewConn(rwc io.ReadWriteCloser, timeout time.Duration) (*Conn, error) {
+	c := &Conn{rwc: rwc, r: bufio.NewReaderSize(rwc, 64<<10), timeout: timeout}
+	line, err := c.readLine()
+	if err != nil {
+		return nil, fmt.Errorf("email: greeting: %w", err)
+	}
+	if !strings.HasPrefix(line, "* OK") && !strings.HasPrefix(line, "* PREAUTH") {
+		c.rwc.Close()
+		return nil, fmt.Errorf("email: server refused connection: %s", line)
+	}
+	return c, nil
+}
+
+func (c *Conn) Close() error { return c.rwc.Close() }
+
+// Login authenticates with LOGIN. The password is passed as an argument and
+// never stored on the Conn: nothing in this package keeps a credential alive
+// past the command that uses it, so no struct dump or panic trace can carry it.
+func (c *Conn) Login(user, pass string) error {
+	// The command line itself is never logged (see exec) — a LOGIN line IS the
+	// credential.
+	if _, err := c.exec(fmt.Sprintf("LOGIN %s %s", quote(user), quote(pass))); err != nil {
+		return fmt.Errorf("email: login: %w", err)
+	}
+	return nil
+}
+
+// Select opens a mailbox read-only. EXAMINE, not SELECT: read-only at the
+// protocol level means no command in this session can change a flag, expunge a
+// message, or move anything, even by mistake.
+func (c *Conn) Select(mailbox string) error {
+	if _, err := c.exec(fmt.Sprintf("EXAMINE %s", quote(mailbox))); err != nil {
+		return fmt.Errorf("email: examine %s: %w", mailbox, err)
+	}
+	return nil
+}
+
+// SearchSince returns the UIDs of messages received on or after since. An
+// unlimited search is not offered: the first poll against a years-old mailbox
+// would otherwise fetch everything and hand a decade of mail to the model.
+//
+// The IMAP SINCE key has date granularity (and compares the server's internal
+// date), so the result can include messages slightly older than since. The
+// caller dedupes by UID anyway, so a wider window costs one extra fetch.
+func (c *Conn) SearchSince(since time.Time) ([]uint32, error) {
+	cmd := fmt.Sprintf("UID SEARCH SINCE %s", since.Format("2-Jan-2006"))
+	lines, err := c.exec(cmd)
+	if err != nil {
+		return nil, fmt.Errorf("email: search: %w", err)
+	}
+	var uids []uint32
+	for _, l := range lines {
+		rest, ok := untagged(l, "SEARCH")
+		if !ok {
+			continue
+		}
+		for _, f := range strings.Fields(rest) {
+			n, err := strconv.ParseUint(f, 10, 32)
+			if err == nil {
+				uids = append(uids, uint32(n))
+			}
+		}
+	}
+	return uids, nil
+}
+
+var literalSize = regexp.MustCompile(`\{(\d+)\}$`)
+
+// Fetch returns the raw RFC 5322 bytes of one message, by UID.
+//
+// Returns (nil, nil) when the UID no longer exists — a message he deleted
+// between SEARCH and FETCH is normal, not an error.
+func (c *Conn) Fetch(uid uint32) ([]byte, error) {
+	tag := c.nextTag()
+	if err := c.send(fmt.Sprintf("%s UID FETCH %d (BODY.PEEK[])", tag, uid)); err != nil {
+		return nil, err
+	}
+	var raw []byte
+	for {
+		line, err := c.readLine()
+		if err != nil {
+			return nil, fmt.Errorf("email: fetch %d: %w", uid, err)
+		}
+		if done, err := c.tagged(tag, line); done {
+			if err != nil {
+				return nil, fmt.Errorf("email: fetch %d: %w", uid, err)
+			}
+			return raw, nil
+		}
+		m := literalSize.FindStringSubmatch(strings.TrimSpace(line))
+		if m == nil {
+			continue
+		}
+		n, err := strconv.Atoi(m[1])
+		if err != nil {
+			continue
+		}
+		buf := make([]byte, n)
+		if _, err := io.ReadFull(c.r, buf); err != nil {
+			return nil, fmt.Errorf("email: fetch %d: literal: %w", uid, err)
+		}
+		if raw == nil {
+			raw = buf
+		}
+	}
+}
+
+// Logout ends the session politely. A failure is not worth reporting — the
+// connection is being closed either way.
+func (c *Conn) Logout() {
+	_, _ = c.exec("LOGOUT")
+}
+
+// ---- protocol plumbing -----------------------------------------------------
+
+func (c *Conn) nextTag() string {
+	c.tag++
+	return fmt.Sprintf("a%03d", c.tag)
+}
+
+// exec sends one command and returns the untagged response lines.
+//
+// Neither the command nor the response is ever logged here. LOGIN goes through
+// this function, and a debug line "sent: a001 LOGIN ..." is how a credential
+// ends up in a log file forever.
+func (c *Conn) exec(cmd string) ([]string, error) {
+	tag := c.nextTag()
+	if err := c.send(tag + " " + cmd); err != nil {
+		return nil, err
+	}
+	var lines []string
+	for {
+		line, err := c.readLine()
+		if err != nil {
+			return nil, err
+		}
+		if done, err := c.tagged(tag, line); done {
+			return lines, err
+		}
+		lines = append(lines, line)
+		// A response line may carry a literal (e.g. a header FETCH). Nothing we
+		// send asks for one outside Fetch, but skip it if it appears so the
+		// stream stays aligned.
+		if m := literalSize.FindStringSubmatch(strings.TrimSpace(line)); m != nil {
+			if n, err := strconv.Atoi(m[1]); err == nil {
+				if _, err := io.CopyN(io.Discard, c.r, int64(n)); err != nil {
+					return nil, err
+				}
+			}
+		}
+	}
+}
+
+// tagged reports whether line completes the command with this tag, and turns a
+// NO/BAD completion into an error. The error text is the server's, which never
+// echoes a password.
+func (c *Conn) tagged(tag, line string) (bool, error) {
+	if !strings.HasPrefix(line, tag+" ") {
+		return false, nil
+	}
+	rest := strings.TrimSpace(line[len(tag):])
+	switch {
+	case strings.HasPrefix(rest, "OK"):
+		return true, nil
+	case strings.HasPrefix(rest, "NO"), strings.HasPrefix(rest, "BAD"):
+		return true, fmt.Errorf("server said: %s", rest)
+	default:
+		return true, fmt.Errorf("unexpected completion: %s", rest)
+	}
+}
+
+func (c *Conn) send(line string) error {
+	c.setDeadline()
+	if _, err := io.WriteString(c.rwc, line+"\r\n"); err != nil {
+		return fmt.Errorf("email: write: %w", err)
+	}
+	return nil
+}
+
+func (c *Conn) readLine() (string, error) {
+	c.setDeadline()
+	line, err := c.r.ReadString('\n')
+	if err != nil {
+		return "", err
+	}
+	return strings.TrimRight(line, "\r\n"), nil
+}
+
+// setDeadline applies the per-connection timeout when the transport supports
+// one. A hung IMAP server must not park the poller forever.
+func (c *Conn) setDeadline() {
+	if c.timeout <= 0 {
+		return
+	}
+	if d, ok := c.rwc.(interface{ SetDeadline(time.Time) error }); ok {
+		_ = d.SetDeadline(time.Now().Add(c.timeout))
+	}
+}
+
+// untagged splits "* SEARCH 1 2 3" into its payload when the key matches.
+func untagged(line, key string) (string, bool) {
+	if !strings.HasPrefix(line, "* ") {
+		return "", false
+	}
+	rest := strings.TrimSpace(line[2:])
+	if !strings.HasPrefix(rest, key) {
+		return "", false
+	}
+	return strings.TrimSpace(rest[len(key):]), true
+}
+
+// quote renders an IMAP quoted string. Passwords routinely contain characters
+// that would otherwise end the argument early, and CR/LF are stripped rather
+// than escaped because there is no legal way to send them — a credential file
+// with a stray newline must not become a second command.
+func quote(s string) string {
+	s = strings.NewReplacer("\r", "", "\n", "").Replace(s)
+	return `"` + strings.NewReplacer(`\`, `\\`, `"`, `\"`).Replace(s) + `"`
+}

internal/email/imap_test.go

@@ -0,0 +1,162 @@
+package email
+
+import (
+	"bufio"
+	"fmt"
+	"net"
+	"strconv"
+	"strings"
+	"testing"
+	"time"
+)
+
+// fakeIMAP is a scripted server: enough of IMAP to exercise the client, and
+// nothing more. It records the commands it received so a test can assert on the
+// protocol (BODY.PEEK rather than BODY, EXAMINE rather than SELECT).
+type fakeIMAP struct {
+	msgs   map[uint32]string
+	uids   []uint32
+	cmds   []string
+	failOn string // substring of a command to answer NO
+}
+
+func (f *fakeIMAP) serve(t *testing.T, c net.Conn) {
+	t.Helper()
+	defer c.Close()
+	fmt.Fprint(c, "* OK fake IMAP ready\r\n")
+	r := bufio.NewReader(c)
+	for {
+		line, err := r.ReadString('\n')
+		if err != nil {
+			return
+		}
+		line = strings.TrimRight(line, "\r\n")
+		parts := strings.SplitN(line, " ", 2)
+		if len(parts) != 2 {
+			return
+		}
+		tag, cmd := parts[0], parts[1]
+		f.cmds = append(f.cmds, cmd)
+		if f.failOn != "" && strings.Contains(cmd, f.failOn) {
+			fmt.Fprintf(c, "%s NO computer says no\r\n", tag)
+			continue
+		}
+		upper := strings.ToUpper(cmd)
+		switch {
+		case strings.HasPrefix(upper, "LOGIN"), strings.HasPrefix(upper, "EXAMINE"):
+			fmt.Fprintf(c, "%s OK done\r\n", tag)
+		case strings.HasPrefix(upper, "UID SEARCH"):
+			var ids []string
+			for _, u := range f.uids {
+				ids = append(ids, strconv.FormatUint(uint64(u), 10))
+			}
+			fmt.Fprintf(c, "* SEARCH %s\r\n", strings.Join(ids, " "))
+			fmt.Fprintf(c, "%s OK search done\r\n", tag)
+		case strings.HasPrefix(upper, "UID FETCH"):
+			uid64, _ := strconv.ParseUint(strings.Fields(cmd)[2], 10, 32)
+			raw, ok := f.msgs[uint32(uid64)]
+			if ok {
+				fmt.Fprintf(c, "* 1 FETCH (UID %d BODY[] {%d}\r\n", uid64, len(raw))
+				fmt.Fprint(c, raw)
+				fmt.Fprint(c, ")\r\n")
+			}
+			fmt.Fprintf(c, "%s OK fetch done\r\n", tag)
+		case strings.HasPrefix(upper, "LOGOUT"):
+			fmt.Fprint(c, "* BYE\r\n")
+			fmt.Fprintf(c, "%s OK bye\r\n", tag)
+			return
+		default:
+			fmt.Fprintf(c, "%s BAD unknown\r\n", tag)
+		}
+	}
+}
+
+// dialFake wires a client Conn to an in-process server over net.Pipe.
+func dialFake(t *testing.T, f *fakeIMAP) *Conn {
+	t.Helper()
+	cli, srv := net.Pipe()
+	go f.serve(t, srv)
+	c, err := NewConn(cli, 5*time.Second)
+	if err != nil {
+		t.Fatalf("greeting: %v", err)
+	}
+	t.Cleanup(func() { c.Close() })
+	return c
+}
+
+func TestIMAPRoundTrip(t *testing.T) {
+	body := "Subject: hello\r\nContent-Type: text/plain; charset=utf-8\r\n\r\nCall the bank.\r\n"
+	f := &fakeIMAP{uids: []uint32{4, 9}, msgs: map[uint32]string{4: body, 9: body}}
+	c := dialFake(t, f)
+
+	if err := c.Login("kami", `pa"ss\word`); err != nil {
+		t.Fatalf("login: %v", err)
+	}
+	if err := c.Select("INBOX"); err != nil {
+		t.Fatalf("select: %v", err)
+	}
+	uids, err := c.SearchSince(time.Date(2026, 8, 1, 0, 0, 0, 0, time.UTC))
+	if err != nil {
+		t.Fatalf("search: %v", err)
+	}
+	if len(uids) != 2 || uids[0] != 4 || uids[1] != 9 {
+		t.Fatalf("uids = %v, want [4 9]", uids)
+	}
+	raw, err := c.Fetch(9)
+	if err != nil {
+		t.Fatalf("fetch: %v", err)
+	}
+	if string(raw) != body {
+		t.Errorf("fetched %q, want the literal verbatim", raw)
+	}
+	c.Logout()
+
+	joined := strings.Join(f.cmds, "\n")
+	// Read-only at the protocol level, and peeking — Maven must leave no trace
+	// of having read his mail.
+	if !strings.Contains(joined, "EXAMINE") || strings.Contains(joined, "SELECT ") {
+		t.Errorf("want EXAMINE (read-only), got:\n%s", joined)
+	}
+	if !strings.Contains(joined, "BODY.PEEK[]") {
+		t.Errorf("want BODY.PEEK, got:\n%s", joined)
+	}
+	// The password must have been quoted and escaped, not truncated at the quote.
+	if !strings.Contains(joined, `"pa\"ss\\word"`) {
+		t.Errorf("password not quoted correctly:\n%s", joined)
+	}
+	// SINCE must carry the IMAP date form.
+	if !strings.Contains(joined, "SINCE 1-Aug-2026") {
+		t.Errorf("want a SINCE date, got:\n%s", joined)
+	}
+}
+
+func TestIMAPServerNoIsAnError(t *testing.T) {
+	f := &fakeIMAP{failOn: "LOGIN"}
+	c := dialFake(t, f)
+	err := c.Login("kami", "wrong")
+	if err == nil {
+		t.Fatal("a NO completion must be an error")
+	}
+	// The error is the server's text; it must not echo the credential.
+	if strings.Contains(err.Error(), "wrong") {
+		t.Errorf("error leaks the password: %v", err)
+	}
+}
+
+func TestIMAPFetchMissingUID(t *testing.T) {
+	f := &fakeIMAP{uids: []uint32{1}, msgs: map[uint32]string{}}
+	c := dialFake(t, f)
+	raw, err := c.Fetch(1)
+	if err != nil {
+		t.Fatalf("fetch: %v", err)
+	}
+	if raw != nil {
+		t.Errorf("a vanished UID should give nil, got %q", raw)
+	}
+}
+
+func TestQuoteStripsNewlines(t *testing.T) {
+	if got := quote("pass\r\nA1 LOGOUT"); strings.ContainsAny(got, "\r\n") {
+		t.Errorf("quote kept a line break: %q", got)
+	}
+}

internal/email/junk.go

@@ -0,0 +1,80 @@
+package email
+
+import (
+	"net/mail"
+	"strings"
+)
+
+// The junk filter — the cheapest and most important half of reading mail.
+//
+// A mailbox is mostly machine-generated: newsletters, receipts nobody acts on,
+// social notifications, marketing. Sending all of it to a 1.7B and asking "is
+// there a task here" produces confident nonsense at a rate proportional to the
+// volume, so junk is decided by HEADERS, before any model sees the message.
+//
+// The rules are all bulk-mail markers that senders set on themselves, never
+// guesses about content:
+//
+//   - List-Unsubscribe / List-Id — by definition a mailing list. If he can
+//     unsubscribe from it, it is not asking him to do anything.
+//   - Precedence: bulk|junk|list — the sender declaring itself bulk.
+//   - Auto-Submitted other than "no" (RFC 3834) — generated by a machine.
+//   - X-Spam-Flag: YES, X-Spam-Status: Yes — the spam filter upstream already
+//     decided; we do not second-guess it in the other direction.
+//   - X-GM-LABELS / X-Gmail-Labels containing a Gmail category — Gmail's own
+//     Promotions/Social/Forums/Spam classification, when the server sends it.
+//
+// Deliberately NOT here: sender allow/deny lists and subject keyword matching.
+// Both are configuration that ages badly and both would be a place for his
+// contacts to end up in a config file. If a real correspondent's mail is being
+// dropped, the fix is a rule about a header, not a list of names.
+//
+// A junk verdict never deletes anything and never touches a flag on the server.
+// It means "do not spend the model on this", nothing more.
+
+// junkHeaders — headers whose mere presence marks bulk mail.
+var junkPresence = []string{"List-Unsubscribe", "List-Id", "List-Post"}
+
+// gmailCategories — Gmail's category labels, lowercased as they appear in
+// X-GM-LABELS. "important" and "inbox" are labels too, and are NOT categories.
+// Matching is by these exact tokens (substring is fine — they are namespaced
+// and cannot appear in a hand-made label by accident), so a user label named
+// "Social Club" is not mistaken for Gmail's Social category.
+var gmailCategories = []string{
+	"category_promotions", "category_social", "category_forums", "category_updates",
+	`\spam`, `\junk`,
+}
+
+// classifyJunk returns whether the message is bulk/automated and why. The
+// reason is a short header name, safe to log — it names the marker, never the
+// sender or the subject.
+func classifyJunk(h mail.Header) (bool, string) {
+	for _, name := range junkPresence {
+		if strings.TrimSpace(h.Get(name)) != "" {
+			return true, strings.ToLower(name)
+		}
+	}
+	switch strings.ToLower(strings.TrimSpace(h.Get("Precedence"))) {
+	case "bulk", "junk", "list":
+		return true, "precedence"
+	}
+	if v := strings.ToLower(strings.TrimSpace(h.Get("Auto-Submitted"))); v != "" && v != "no" {
+		return true, "auto-submitted"
+	}
+	if strings.EqualFold(strings.TrimSpace(h.Get("X-Spam-Flag")), "yes") {
+		return true, "x-spam-flag"
+	}
+	if v := strings.ToLower(strings.TrimSpace(h.Get("X-Spam-Status"))); strings.HasPrefix(v, "yes") {
+		return true, "x-spam-status"
+	}
+	labels := strings.ToLower(h.Get("X-GM-LABELS") + " " + h.Get("X-Gmail-Labels"))
+	for _, c := range gmailCategories {
+		if c == "" {
+			continue
+		}
+		if strings.Contains(labels, c) {
+			return true, "gmail-category"
+		}
+	}
+	return false, ""
+}

internal/email/junk_test.go

@@ -0,0 +1,59 @@
+package email
+
+import (
+	"net/mail"
+	"strings"
+	"testing"
+)
+
+func headers(t *testing.T, raw string) mail.Header {
+	t.Helper()
+	m, err := mail.ReadMessage(strings.NewReader(strings.ReplaceAll(raw, "\n", "\r\n") + "\r\n\r\nbody\r\n"))
+	if err != nil {
+		t.Fatalf("read headers: %v", err)
+	}
+	return m.Header
+}
+
+func TestClassifyJunk(t *testing.T) {
+	cases := []struct {
+		name   string
+		raw    string
+		junk   bool
+		reason string
+	}{
+		{"personal", "From: a@b.c\nSubject: привет", false, ""},
+		{"list-unsubscribe", "From: a@b.c\nList-Unsubscribe: <mailto:u@b.c>", true, "list-unsubscribe"},
+		{"list-id", "From: a@b.c\nList-Id: <golang-nuts.example>", true, "list-id"},
+		{"precedence bulk", "From: a@b.c\nPrecedence: bulk", true, "precedence"},
+		{"auto-submitted", "From: a@b.c\nAuto-Submitted: auto-generated", true, "auto-submitted"},
+		{"auto-submitted no", "From: a@b.c\nAuto-Submitted: no", false, ""},
+		{"spam flag", "From: a@b.c\nX-Spam-Flag: YES", true, "x-spam-flag"},
+		{"spam status", "From: a@b.c\nX-Spam-Status: Yes, score=9.1", true, "x-spam-status"},
+		{"spam status no", "From: a@b.c\nX-Spam-Status: No, score=0.1", false, ""},
+		{"gmail promo", "From: a@b.c\nX-Gmail-Labels: Inbox,CATEGORY_PROMOTIONS", true, "gmail-category"},
+		{"user label", "From: a@b.c\nX-Gmail-Labels: Social Club,Important", false, ""},
+	}
+	for _, c := range cases {
+		t.Run(c.name, func(t *testing.T) {
+			junk, reason := classifyJunk(headers(t, c.raw))
+			if junk != c.junk || reason != c.reason {
+				t.Errorf("classifyJunk = (%v, %q), want (%v, %q)", junk, reason, c.junk, c.reason)
+			}
+		})
+	}
+}
+
+func TestNewsletterFixtureIsJunk(t *testing.T) {
+	msg, err := ParseMessage(9, fixture(t, "newsletter.eml"))
+	if err != nil {
+		t.Fatalf("parse: %v", err)
+	}
+	if !msg.Junk {
+		t.Fatal("a newsletter with List-Unsubscribe + Precedence: bulk must be junk")
+	}
+	// The reason is what gets logged, so it must never carry mail content.
+	if strings.Contains(msg.JunkReason, "@") || strings.Contains(msg.JunkReason, "Скидки") {
+		t.Errorf("junk reason leaks content: %q", msg.JunkReason)
+	}
+}

internal/email/message.go

@@ -0,0 +1,258 @@
+// Package email is the reading half of the email reader (Vikunja #246,
+// docs/plans/01-email-reader.md): a small IMAP client, a MIME-to-plaintext
+// converter, and the junk filter that decides a message is not worth reading at
+// all. Extraction lives in extract.go and writes nothing itself.
+//
+// Two constraints shape everything here, both from CLAUDE.md:
+//
+//   - Mail is personal. Nothing in this package logs a body, a subject, or an
+//     address; callers get the text and decide. Mail text is never search input
+//     — no function here reaches the network except the IMAP connection itself.
+//   - Off unless configured. There is no default host, no default account, and
+//     no fallback that would make a mailbox get read because a field was empty.
+//
+// The IMAP subset is deliberately tiny (LOGIN, SELECT, UID SEARCH, UID FETCH
+// with BODY.PEEK, LOGOUT). No IDLE: a poll every few minutes is what a task
+// candidate needs, and IDLE would mean holding a connection and a credential
+// open forever for latency nobody is waiting on.
+package email
+
+import (
+	"encoding/base64"
+	"fmt"
+	"io"
+	"mime"
+	"mime/multipart"
+	"mime/quotedprintable"
+	"net/mail"
+	"regexp"
+	"strings"
+)
+
+// MaxBodyBytes — how much of one message body is kept. A task hides in the
+// first screenful; the rest is signature, quoted history and legal boilerplate,
+// and it would only spend the resident model's 4096-token context.
+const MaxBodyBytes = 4000
+
+// Message — one mail, reduced to the fields extraction and review need.
+//
+// Raw is deliberately absent: once a message is parsed the original bytes are
+// dropped, so no caller can accidentally log or forward the whole mail.
+type Message struct {
+	UID     uint32
+	From    string
+	Subject string
+	Date    string // as sent, unparsed — display only
+	Body    string // plaintext, decoded, HTML-stripped, truncated
+	// Junk is set by the junk filter (see junk.go). A junk message is carried
+	// rather than dropped so the poller can count it and still mark it seen.
+	Junk       bool
+	JunkReason string
+}
+
+// ParseMessage turns one RFC 5322 message into a Message.
+//
+// It never fails on a body it cannot understand: an unparsable or
+// unsupported-charset body yields an empty Body and the headers still come
+// through, because a subject line alone is often the whole task ("Счёт за
+// интернет"). Only a message whose headers cannot be read at all is an error.
+func ParseMessage(uid uint32, raw []byte) (Message, error) {
+	m, err := mail.ReadMessage(strings.NewReader(string(raw)))
+	if err != nil {
+		return Message{}, fmt.Errorf("email: parse message: %w", err)
+	}
+	msg := Message{
+		UID:     uid,
+		From:    decodeHeader(m.Header.Get("From")),
+		Subject: decodeHeader(m.Header.Get("Subject")),
+		Date:    m.Header.Get("Date"),
+	}
+	msg.Junk, msg.JunkReason = classifyJunk(m.Header)
+	body, err := plaintextBody(m.Header.Get("Content-Type"), m.Header.Get("Content-Transfer-Encoding"), m.Body)
+	if err == nil {
+		msg.Body = truncate(collapse(body), MaxBodyBytes)
+	}
+	return msg, nil
+}
+
+// plaintextBody walks the MIME tree and returns the best plaintext it can.
+//
+// Preference order inside a multipart: text/plain first, text/html stripped
+// only when there is no plain part. multipart/mixed attachments are skipped
+// wholesale — an attachment is a file, not a sentence, and reading one would
+// mean parsing arbitrary formats from the network.
+func plaintextBody(contentType, encoding string, body io.Reader) (string, error) {
+	mediaType, params, err := mime.ParseMediaType(contentType)
+	if contentType == "" || err != nil {
+		// No Content-Type at all is legal and means text/plain; a broken one is
+		// treated the same rather than dropping the message.
+		mediaType, params = "text/plain", nil
+	}
+	switch {
+	case strings.HasPrefix(mediaType, "multipart/"):
+		boundary := params["boundary"]
+		if boundary == "" {
+			return "", fmt.Errorf("email: multipart without boundary")
+		}
+		return multipartText(multipart.NewReader(body, boundary))
+	case mediaType == "text/html":
+		raw, err := decodeBody(body, encoding, params["charset"])
+		if err != nil {
+			return "", err
+		}
+		return stripHTML(raw), nil
+	case mediaType == "text/plain":
+		return decodeBody(body, encoding, params["charset"])
+	default:
+		// A single-part non-text message (a bare PDF, say). No body, headers only.
+		return "", nil
+	}
+}
+
+// multipartText reads one multipart level, recursing into nested multiparts.
+// Returns the plain part if any part yielded one, else the stripped HTML.
+func multipartText(mr *multipart.Reader) (string, error) {
+	var plain, html string
+	for {
+		part, err := mr.NextPart()
+		if err == io.EOF {
+			break
+		}
+		if err != nil {
+			// A truncated multipart still gives up whatever came before it.
+			break
+		}
+		if part.FileName() != "" {
+			part.Close()
+			continue // attachment
+		}
+		ct := part.Header.Get("Content-Type")
+		mediaType, _, _ := mime.ParseMediaType(ct)
+		text, err := plaintextBody(ct, part.Header.Get("Content-Transfer-Encoding"), part)
+		part.Close()
+		if err != nil || strings.TrimSpace(text) == "" {
+			continue
+		}
+		if mediaType == "text/html" && !strings.HasPrefix(mediaType, "multipart/") {
+			if html == "" {
+				html = text
+			}
+			continue
+		}
+		if plain == "" {
+			plain = text
+		}
+	}
+	if strings.TrimSpace(plain) != "" {
+		return plain, nil
+	}
+	return html, nil
+}
+
+// decodeBody applies the transfer encoding, then the charset.
+//
+// Charset support is UTF-8 (and ASCII, its subset) only, on purpose: x/text's
+// encoding tables are not vendored here, and guessing at windows-1251 bytes
+// would feed the model mojibake it would happily extract a task from. An
+// unsupported charset returns an error, which ParseMessage turns into an empty
+// body — subject-only, which is honest.
+func decodeBody(r io.Reader, encoding, charset string) (string, error) {
+	switch strings.ToLower(strings.TrimSpace(encoding)) {
+	case "quoted-printable":
+		r = quotedprintable.NewReader(r)
+	case "base64":
+		r = newBase64Reader(r)
+	}
+	b, err := io.ReadAll(io.LimitReader(r, 1<<20))
+	if err != nil && len(b) == 0 {
+		return "", fmt.Errorf("email: read body: %w", err)
+	}
+	switch cs := strings.ToLower(strings.TrimSpace(charset)); cs {
+	case "", "utf-8", "utf8", "us-ascii", "ascii":
+		return string(b), nil
+	default:
+		return "", fmt.Errorf("email: unsupported charset %q", cs)
+	}
+}
+
+// decodeHeader decodes RFC 2047 encoded words ("=?utf-8?B?...?="), which is how
+// every Russian subject line arrives. Undecodable headers come back as-is
+// rather than empty: a mangled subject is still a hint, and it is only ever
+// shown to him as evidence.
+func decodeHeader(v string) string {
+	dec := new(mime.WordDecoder)
+	out, err := dec.DecodeHeader(v)
+	if err != nil {
+		return collapse(v)
+	}
+	return collapse(out)
+}
+
+var (
+	scriptStyle = regexp.MustCompile(`(?is)<(script|style)\b[^>]*>.*?</\s*(script|style)\s*>`)
+	htmlBreak   = regexp.MustCompile(`(?i)<\s*(br\s*/?|/p|/div|/tr|/li|/h[1-6])\s*>`)
+	htmlTag     = regexp.MustCompile(`(?s)<[^>]*>`)
+	htmlComment = regexp.MustCompile(`(?s)<!--.*?-->`)
+)
+
+// stripHTML reduces an HTML part to text. A regex stripper, not a parser:
+// x/net/html is not vendored, and the consumer is a model reading prose — a
+// stray angle bracket costs nothing, whereas a new dependency for the privacy-
+// sensitive path costs review.
+func stripHTML(s string) string {
+	s = scriptStyle.ReplaceAllString(s, " ")
+	s = htmlComment.ReplaceAllString(s, " ")
+	s = htmlBreak.ReplaceAllString(s, "\n")
+	s = htmlTag.ReplaceAllString(s, " ")
+	return unescapeEntities(s)
+}
+
+var entities = strings.NewReplacer(
+	"&nbsp;", " ", "&amp;", "&", "&lt;", "<", "&gt;", ">",
+	"&quot;", `"`, "&#39;", "'", "&apos;", "'", "&mdash;", "—", "&ndash;", "–",
+)
+
+func unescapeEntities(s string) string { return entities.Replace(s) }
+
+// collapse squeezes runs of whitespace, keeping single newlines. Mail bodies
+// arrive with hard-wrapped lines and blocks of blank space; the model does not
+// need them and they are pure context budget.
+func collapse(s string) string {
+	lines := strings.Split(strings.ReplaceAll(s, "\r\n", "\n"), "\n")
+	var out []string
+	blank := 0
+	for _, l := range lines {
+		l = strings.TrimSpace(strings.Join(strings.Fields(l), " "))
+		if l == "" {
+			blank++
+			if blank > 1 {
+				continue
+			}
+			out = append(out, "")
+			continue
+		}
+		blank = 0
+		out = append(out, l)
+	}
+	return strings.TrimSpace(strings.Join(out, "\n"))
+}
+
+// truncate cuts to n bytes on a rune boundary.
+func truncate(s string, n int) string {
+	if len(s) <= n {
+		return s
+	}
+	cut := s[:n]
+	for len(cut) > 0 && !isRuneStart(cut[len(cut)-1]) {
+		cut = cut[:len(cut)-1]
+	}
+	return strings.TrimSpace(cut) + "…"
+}
+
+func isRuneStart(b byte) bool { return b&0xC0 != 0x80 }
+
+// newBase64Reader — base64.NewDecoder already skips the CRLFs mail bodies wrap
+// with, so this is only a named seam for decodeBody to read cleanly.
+func newBase64Reader(r io.Reader) io.Reader {
+	return base64.NewDecoder(base64.StdEncoding, r)
+}

internal/email/message_test.go

@@ -0,0 +1,111 @@
+package email
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+func fixture(t *testing.T, name string) []byte {
+	t.Helper()
+	b, err := os.ReadFile(filepath.Join("testdata", name))
+	if err != nil {
+		t.Fatalf("read fixture %s: %v", name, err)
+	}
+	return b
+}
+
+func TestParsePlainRussian(t *testing.T) {
+	msg, err := ParseMessage(7, fixture(t, "plain_ru.eml"))
+	if err != nil {
+		t.Fatalf("parse: %v", err)
+	}
+	if msg.UID != 7 {
+		t.Errorf("uid = %d, want 7", msg.UID)
+	}
+	if want := "Нужно закрыть задачу"; msg.Subject != want {
+		t.Errorf("subject = %q, want %q", msg.Subject, want)
+	}
+	if !strings.Contains(msg.From, "Антон") {
+		t.Errorf("from = %q, want the decoded display name", msg.From)
+	}
+	if !strings.Contains(msg.Body, "Надо отправить акт до пятницы.") {
+		t.Errorf("body = %q, want the quoted-printable text decoded", msg.Body)
+	}
+	if msg.Junk {
+		t.Errorf("a personal mail must not be junk (%s)", msg.JunkReason)
+	}
+}
+
+func TestParseHTMLOnlyIsStripped(t *testing.T) {
+	msg, err := ParseMessage(1, fixture(t, "html_only.eml"))
+	if err != nil {
+		t.Fatalf("parse: %v", err)
+	}
+	if strings.Contains(msg.Body, "<") || strings.Contains(msg.Body, "color:red") || strings.Contains(msg.Body, "x()") {
+		t.Errorf("body still has markup/script/style: %q", msg.Body)
+	}
+	for _, want := range []string{"Счёт за интернет: 700", "Оплатить до 5 августа."} {
+		if !strings.Contains(msg.Body, want) {
+			t.Errorf("body = %q, want it to contain %q", msg.Body, want)
+		}
+	}
+	// &nbsp; must have become a real space, not vanished into the number.
+	if strings.Contains(msg.Body, "&nbsp;") {
+		t.Errorf("entity left unescaped: %q", msg.Body)
+	}
+}
+
+func TestParsePrefersPlainAndSkipsAttachments(t *testing.T) {
+	msg, err := ParseMessage(2, fixture(t, "mixed_attachment.eml"))
+	if err != nil {
+		t.Fatalf("parse: %v", err)
+	}
+	if got := strings.TrimSpace(msg.Body); got != "Sign the contract before Monday." {
+		t.Errorf("body = %q, want the text/plain alternative only", got)
+	}
+	if strings.Contains(msg.Body, "PDF") {
+		t.Errorf("attachment bytes leaked into the body: %q", msg.Body)
+	}
+}
+
+// An unsupported charset must degrade to headers-only rather than to mojibake
+// the model would then extract a task from.
+func TestParseUnsupportedCharsetKeepsHeaders(t *testing.T) {
+	msg, err := ParseMessage(3, fixture(t, "cp1251.eml"))
+	if err != nil {
+		t.Fatalf("parse: %v", err)
+	}
+	if msg.Subject != "Legacy" {
+		t.Errorf("subject = %q, want Legacy", msg.Subject)
+	}
+	if msg.Body != "" {
+		t.Errorf("body = %q, want empty for an undecodable charset", msg.Body)
+	}
+}
+
+func TestParseTruncatesLongBody(t *testing.T) {
+	var b strings.Builder
+	b.WriteString("Subject: long\r\nContent-Type: text/plain; charset=utf-8\r\n\r\n")
+	for i := 0; i < 2000; i++ {
+		b.WriteString("длинная строка ")
+	}
+	msg, err := ParseMessage(4, []byte(b.String()))
+	if err != nil {
+		t.Fatalf("parse: %v", err)
+	}
+	if len(msg.Body) > MaxBodyBytes+8 {
+		t.Errorf("body kept %d bytes, want ≤ %d", len(msg.Body), MaxBodyBytes)
+	}
+	if !strings.HasSuffix(msg.Body, "…") {
+		t.Errorf("truncated body should be marked: %q", msg.Body[len(msg.Body)-20:])
+	}
+}
+
+func TestCollapseSqueezesBlankLines(t *testing.T) {
+	got := collapse("  a  b \r\n\r\n\r\n\r\n  c  \r\n")
+	if got != "a b\n\nc" {
+		t.Errorf("collapse = %q, want %q", got, "a b\n\nc")
+	}
+}

internal/email/testdata/cp1251.eml

@@ -0,0 +1,7 @@
+From: legacy@example.org
+To: kami@example.org
+Subject: Legacy
+Date: Fri, 01 Aug 2026 05:00:00 +0400
+Content-Type: text/plain; charset="windows-1251"
+
+Ï

internal/email/testdata/html_only.eml

@@ -0,0 +1,13 @@
+From: billing@isp.example
+To: kami@example.org
+Subject: =?utf-8?B?0KHRh9GR0YIg0LfQsCDQuNC90YLQtdGA0L3QtdGC?=
+Date: Fri, 01 Aug 2026 08:00:00 +0400
+MIME-Version: 1.0
+Content-Type: multipart/alternative; boundary="B1"
+
+--B1
+Content-Type: text/html; charset="utf-8"
+Content-Transfer-Encoding: base64
+
+PGh0bWw+PGhlYWQ+PHN0eWxlPnB7Y29sb3I6cmVkfTwvc3R5bGU+PC9oZWFkPjxib2R5PjxwPtCh0YfRkdGCINC30LAg0LjQvdGC0LXRgNC90LXRgjogNzAwJm5ic3A74oK9PC9wPjxwPtCe0L/Qu9Cw0YLQuNGC0Ywg0LTQviA1INCw0LLQs9GD0YHRgtCwLjwvcD48c2NyaXB0PngoKTwvc2NyaXB0PjwvYm9keT48L2h0bWw+
+--B1--

internal/email/testdata/mixed_attachment.eml

@@ -0,0 +1,26 @@
+From: hr@work.example
+To: kami@example.org
+Subject: Contract
+Date: Fri, 01 Aug 2026 07:00:00 +0400
+MIME-Version: 1.0
+Content-Type: multipart/mixed; boundary="M1"
+
+--M1
+Content-Type: multipart/alternative; boundary="A1"
+
+--A1
+Content-Type: text/plain; charset="utf-8"
+
+Sign the contract before Monday.
+--A1
+Content-Type: text/html; charset="utf-8"
+
+<p>Sign the contract before Monday.</p>
+--A1--
+--M1
+Content-Type: application/pdf; name="contract.pdf"
+Content-Disposition: attachment; filename="contract.pdf"
+Content-Transfer-Encoding: base64
+
+JVBERi0xLjQgbm90IHJlYWxseSBhIHBkZg==
+--M1--

internal/email/testdata/newsletter.eml

@@ -0,0 +1,9 @@
+From: news@shop.example
+To: kami@example.org
+Subject: =?utf-8?B?0KHQutC40LTQutC4INGC0L7Qu9GM0LrQviDRgdC10LPQvtC00L3Rjw==?=
+Date: Fri, 01 Aug 2026 06:00:00 +0400
+List-Unsubscribe: <mailto:unsub@shop.example>
+Precedence: bulk
+Content-Type: text/plain; charset="utf-8"
+
+Sale!

internal/email/testdata/plain_ru.eml

@@ -0,0 +1,14 @@
+From: =?utf-8?B?0JDQvdGC0L7QvQ==?= <anton@example.org>
+To: kami@example.org
+Subject: =?utf-8?B?0J3Rg9C20L3QviDQt9Cw0LrRgNGL0YLRjCDQt9Cw0LTQsNGH0YM=?=
+Date: Fri, 01 Aug 2026 09:12:00 +0400
+Content-Type: text/plain; charset="utf-8"
+Content-Transfer-Encoding: quoted-printable
+Message-ID: <plain-ru@example.org>
+
+=D0=9F=D1=80=D0=B8=D0=B2=D0=B5=D1=82! =D0=9D=D0=B0=D0=B4=D0=BE =D0=BE=D1=82=
+=D0=BF=D1=80=D0=B0=D0=B2=D0=B8=D1=82=D1=8C =D0=B0=D0=BA=D1=82 =D0=B4=D0=BE =
+=D0=BF=D1=8F=D1=82=D0=BD=D0=B8=D1=86=D1=8B.
+
+--
+Anton

internal/ipc/api.go

@@ -138,6 +138,44 @@ type CaptureTaskResp struct {
 	Created bool  `json:"created"`
 }
 
+// IngestMailReq — one message a mail reader has fetched, handed to core for
+// extraction (Vikunja #246).
+//
+// The mail reader (cmd/mavmaild) holds the IMAP credential and core never sees
+// it, the same split mavpoll uses for the zenmoney token. What crosses this
+// boundary is only the message text, because extraction runs on the resident
+// model and llama-server lives inside core's process.
+//
+// Body is already plaintext and truncated by internal/email; core does not
+// re-parse MIME and never stores the body. Junk means the reader's header
+// filter already classified the message as bulk — core is told rather than
+// asked, so a junk message can be counted without a model call.
+//
+// This method is available only when core has an email block configured AND a
+// llama-server phraser; otherwise it answers ErrUnknownMethod, which is what
+// "off unless configured" looks like at the wire.
+type IngestMailReq struct {
+	Mailbox string `json:"mailbox"`
+	UID     uint32 `json:"uid"`
+	From    string `json:"from,omitempty"`
+	Subject string `json:"subject,omitempty"`
+	Date    string `json:"date,omitempty"`
+	Body    string `json:"body,omitempty"`
+	Junk    bool   `json:"junk,omitempty"`
+}
+
+// IngestMailResp — what core did with the message. TaskIDs are the rows
+// CaptureTask returned; Created counts the ones that were new (a re-read
+// mailbox dedupes to Created=0). Skipped is set when nothing was asked of the
+// model at all — junk, or an empty message.
+//
+// Nothing here echoes the mail back. The reader logs counts.
+type IngestMailResp struct {
+	TaskIDs []int64 `json:"task_ids,omitempty"`
+	Created int     `json:"created"`
+	Skipped bool    `json:"skipped,omitempty"`
+}
+
 type listTasksReq struct {
 	Status string `json:"status"` // "" all | "live" | candidate|open|done|dropped
 }

internal/ipc/client.go

@@ -448,6 +448,17 @@ func (c *Client) SetTaskStatus(ctx context.Context, id int64, status string, ts
 	return c.call(ctx, MethodSetTaskStatus, setTaskStatusReq{ID: id, Status: status, Ts: ts}, nil)
 }
 
+// IngestMail hands one fetched message to core for extraction. ErrUnknownMethod
+// means core has no email block configured — the caller should stop asking, not
+// retry.
+func (c *Client) IngestMail(ctx context.Context, req IngestMailReq) (IngestMailResp, error) {
+	var r IngestMailResp
+	if err := c.call(ctx, MethodIngestMail, req, &r); err != nil {
+		return IngestMailResp{}, err
+	}
+	return r, nil
+}
+
 func (c *Client) DismissProposedRoutine(ctx context.Context, id int64) error {
 	return c.call(ctx, MethodDismissProposedRoutine, dismissProposedRoutineReq{ID: id}, nil)
 }

internal/ipc/ipc_test.go

@@ -565,3 +565,36 @@ func mustJSON(v any) []byte {
 	}
 	return b
 }
+
+// TestIngestMail_OffUnlessConfigured — with no IngestMailFn set (the default,
+// and what an unconfigured core looks like) the method does not exist. A mail
+// reader gets a refusal it can act on rather than a silent success.
+func TestIngestMail_OffUnlessConfigured(t *testing.T) {
+	_, _, cli, _ := newServerWithStore(t)
+	if _, err := cli.IngestMail(context.Background(), IngestMailReq{Mailbox: "INBOX", UID: 1}); !errors.Is(err, ErrUnknownMethod) {
+		t.Fatalf("IngestMail error = %v, want ErrUnknownMethod", err)
+	}
+}
+
+// TestIngestMail_Hook — when the daemon wires the hook, the message crosses the
+// boundary intact and the response comes back.
+func TestIngestMail_Hook(t *testing.T) {
+	_, srv, cli, _ := newServerWithStore(t)
+	var got IngestMailReq
+	srv.IngestMailFn = func(_ context.Context, req IngestMailReq) (IngestMailResp, error) {
+		got = req
+		return IngestMailResp{TaskIDs: []int64{7}, Created: 1}, nil
+	}
+	resp, err := cli.IngestMail(context.Background(), IngestMailReq{
+		Mailbox: "INBOX", UID: 12, Subject: "Счёт", Body: "Оплатить.", Junk: false,
+	})
+	if err != nil {
+		t.Fatalf("IngestMail: %v", err)
+	}
+	if resp.Created != 1 || len(resp.TaskIDs) != 1 || resp.TaskIDs[0] != 7 {
+		t.Errorf("resp = %+v", resp)
+	}
+	if got.UID != 12 || got.Subject != "Счёт" || got.Body != "Оплатить." {
+		t.Errorf("req across the wire = %+v", got)
+	}
+}

internal/ipc/server.go

@@ -421,6 +421,17 @@ type Server struct {
 	// Set by the daemon; nil ⇒ MethodStoreEncryptionKey returns ErrUnknownMethod.
 	WrapKeyFn WrapKeyFunc
 
+	// IngestMailFn — extracts task candidates from one fetched message. Set by
+	// the daemon only when an email block is configured AND there is a
+	// llama-server to extract with; nil ⇒ MethodIngestMail returns
+	// ErrUnknownMethod, so a mail reader pointed at a core that is not
+	// configured for mail is refused rather than silently ignored.
+	//
+	// Like StepUp/WrapKeyFn/UnlockFn this bypasses CoreAPI: it is not a store
+	// operation, it needs the resident model, and it must not become a method
+	// every CoreAPI implementation has to carry.
+	IngestMailFn IngestMailFunc
+
 	// UnlockFn — unwraps the store encryption key from the wrapped blob using
 	// the passkey credential public key, opens the encrypted store, and wires
 	// the rest of the daemon (voice, loop, delivery). Set by the daemon when
@@ -439,6 +450,9 @@ type WrapKeyFunc func(ctx context.Context, publicKey []byte) error
 // public key and completes daemon initialization.
 type UnlockFunc func(ctx context.Context, publicKey []byte) error
 
+// IngestMailFunc — core-side mail extraction. Returns what was captured.
+type IngestMailFunc func(ctx context.Context, req IngestMailReq) (IngestMailResp, error)
+
 // CheckFunc — the auth hook signature. Wired by the daemon (auth.Gate.Check
 // satisfies this); dispatch calls it once per request after param-unmarshal
 // independence (it gets the raw params, may unmarshal what it needs — ipc
@@ -606,9 +620,10 @@ func withoutParams[R any](fn func(ctx context.Context, api CoreAPI) (R, error))
 // existed) as an argument — so SetAPI's runtime swap (the unlock transition)
 // is still honored on the very next request with no extra plumbing here.
 //
-// MethodAssertStepUp, MethodStoreEncryptionKey and MethodUnlock are NOT in
-// this table: they bypass CoreAPI entirely (s.StepUp / s.WrapKeyFn /
-// s.UnlockFn), so dispatch special-cases them before consulting the table.
+// MethodAssertStepUp, MethodStoreEncryptionKey, MethodUnlock and
+// MethodIngestMail are NOT in this table: they bypass CoreAPI entirely
+// (s.StepUp / s.WrapKeyFn / s.UnlockFn / s.IngestMailFn), so dispatch
+// special-cases them before consulting the table.
 var methodTable = map[Method]handlerFunc{
 	MethodWriteFact: withParams(func(ctx context.Context, api CoreAPI, p WriteFactReq) (idResp, error) {
 		id, err := api.WriteFact(ctx, p)
@@ -816,7 +831,7 @@ func (s *Server) dispatch(ctx context.Context, req Request) (json.RawMessage, er
 		}
 	}
 
-	// These three bypass CoreAPI entirely — they drive Server fields set
+	// These bypass CoreAPI entirely — they drive Server fields set
 	// directly by the daemon (StepUp / WrapKeyFn / UnlockFn), not store
 	// state, so they can never be table entries keyed on a CoreAPI method.
 	switch req.Method {
@@ -845,6 +860,20 @@ func (s *Server) dispatch(ctx context.Context, req Request) (json.RawMessage, er
 			return marshalResult(nil), s.UnlockFn(ctx, p.PublicKey)
 		}
 		return nil, fmt.Errorf("%w: %s", ErrUnknownMethod, req.Method)
+
+	case MethodIngestMail:
+		if s.IngestMailFn != nil {
+			var p IngestMailReq
+			if err := unmarshalParams(req.Params, &p); err != nil {
+				return nil, err
+			}
+			resp, err := s.IngestMailFn(ctx, p)
+			if err != nil {
+				return nil, err
+			}
+			return marshalResult(resp), nil
+		}
+		return nil, fmt.Errorf("%w: %s", ErrUnknownMethod, req.Method)
 	}
 
 	h, ok := methodTable[req.Method]

internal/ipc/wire.go

@@ -50,6 +50,7 @@ const (
 	MethodCaptureTask            Method = "capture_task"
 	MethodListTasks              Method = "list_tasks"
 	MethodSetTaskStatus          Method = "set_task_status"
+	MethodIngestMail             Method = "ingest_mail"
 )
 
 // Request — one frame from module to core. Params is the JSON-encoded argument

internal/router/feeds.go

@@ -0,0 +1,109 @@
+package router
+
+import "strings"
+
+// Feed queries — "что нового в лентах?", "что нового по технологиям?"
+// (Vikunja #258).
+//
+// Deterministic matching, like the calendar, plan and habit matchers above it:
+// the LLM router says this is a query, and this decides whether it is a question
+// about the feeds. A model deciding that would occasionally answer "что нового?"
+// out of world knowledge, which is the one thing a feed reader exists to avoid.
+
+// FeedQuery — a parsed "what's new" question. Category is the topic he named
+// ("технологии"), empty when he asked about the feeds in general.
+type FeedQuery struct {
+	Category string
+}
+
+// feedNouns — the words that make a question about the feeds themselves.
+var feedNouns = []string{
+	"лента", "ленте", "ленты", "лентах", "лентам",
+	"новости", "новостей", "новостях", "новостям",
+	"новое", "нового", "новенького",
+	"feed", "feeds", "news", "headlines",
+}
+
+// newnessMarkers — the "что нового" half. "нового" alone is in feedNouns
+// because it carries the question on its own ("что нового?"); a bare "лента"
+// needs the ask, which is what askMarkers below is for.
+var askMarkers = []string{
+	"что", "какие", "какое", "расскажи", "почитай", "прочитай", "покажи",
+	"what", "any", "tell", "show", "read",
+}
+
+// ParseFeedQuery reports whether an utterance asks what is new in the feeds, and
+// which topic if it names one after "по"/"о"/"про"/"about".
+//
+// Both a feed noun and an ask are required. "у меня новая лента в инстаграме" is
+// a statement and must not be read as a request to recite headlines.
+func ParseFeedQuery(text string) (FeedQuery, bool) {
+	toks := planTokens(text)
+	noun, ask := false, false
+	for _, t := range toks {
+		for _, n := range feedNouns {
+			if t == n {
+				noun = true
+				break
+			}
+		}
+		for _, a := range askMarkers {
+			if t == a {
+				ask = true
+				break
+			}
+		}
+	}
+	if !noun || !ask {
+		return FeedQuery{}, false
+	}
+	return FeedQuery{Category: feedCategory(toks)}, true
+}
+
+// categoryPreps — the prepositions a topic follows. Russian marks the topic with
+// a preposition ("по технологиям", "про политику"), so the word after one is the
+// category; there is no stemming here, and the match against the configured
+// category is a prefix comparison for exactly that reason.
+var categoryPreps = map[string]bool{"по": true, "о": true, "об": true, "про": true, "about": true, "on": true}
+
+func feedCategory(toks []string) string {
+	for i, t := range toks {
+		if categoryPreps[t] && i+1 < len(toks) {
+			next := toks[i+1]
+			// "по новостям" names no topic, it repeats the noun.
+			for _, n := range feedNouns {
+				if next == n {
+					return ""
+				}
+			}
+			return next
+		}
+	}
+	return ""
+}
+
+// CategoryMatches reports whether a note's text plausibly belongs to the
+// category he named. Russian inflects the topic ("технологиям" vs the configured
+// "технологии"), and there is no stemmer in this repo, so the comparison is on a
+// common prefix — long enough that "полит" and "погод" stay apart, short enough
+// to survive a case ending.
+func CategoryMatches(text, category string) bool {
+	if category == "" {
+		return true
+	}
+	stem := categoryStem(category)
+	if stem == "" {
+		return false
+	}
+	return strings.Contains(strings.ToLower(text), stem)
+}
+
+// categoryStem cuts a word down to the part inflection leaves alone. 5 runes is
+// the compromise: shorter words are used whole.
+func categoryStem(word string) string {
+	r := []rune(strings.ToLower(strings.TrimSpace(word)))
+	if len(r) > 5 {
+		r = r[:5]
+	}
+	return string(r)
+}

internal/router/feeds_test.go

@@ -0,0 +1,48 @@
+package router
+
+import "testing"
+
+func TestParseFeedQuery(t *testing.T) {
+	cases := []struct {
+		text     string
+		ok       bool
+		category string
+	}{
+		{"что нового в лентах?", true, ""},
+		{"что нового?", true, ""},
+		{"какие новости?", true, ""},
+		{"что нового по технологиям?", true, "технологиям"},
+		{"расскажи новости про политику", true, "политику"},
+		{"что нового по новостям", true, ""},
+		{"what's new in the feeds?", true, ""},
+		{"any news about kubernetes", true, "kubernetes"},
+		// Statements, not requests.
+		{"у меня новая лента в инстаграме", false, ""},
+		{"новости меня утомили", false, ""},
+		{"напомни полить цветы", false, ""},
+		{"", false, ""},
+	}
+	for _, c := range cases {
+		q, ok := ParseFeedQuery(c.text)
+		if ok != c.ok {
+			t.Errorf("ParseFeedQuery(%q) ok = %v, want %v", c.text, ok, c.ok)
+			continue
+		}
+		if ok && q.Category != c.category {
+			t.Errorf("ParseFeedQuery(%q) category = %q, want %q", c.text, q.Category, c.category)
+		}
+	}
+}
+
+func TestCategoryMatches(t *testing.T) {
+	// The inflected form he says must match the form the config spells.
+	if !CategoryMatches("Новый релиз [технологии]", "технологиям") {
+		t.Error("inflected category did not match")
+	}
+	if CategoryMatches("Новый релиз [технологии]", "политику") {
+		t.Error("unrelated category matched")
+	}
+	if !CategoryMatches("anything", "") {
+		t.Error("an empty category must match everything")
+	}
+}

internal/router/money.go

@@ -0,0 +1,71 @@
+package router
+
+import "strings"
+
+// Money questions, matched deterministically (Vikunja #125).
+//
+// No new intent, for the same reason as tasks: the intent enum is a contract
+// with the relabelling prompt. "сколько я потратил?" is a query; which figure
+// it asks for is a lookup, not something to ask a 1.7B — and a model asked to
+// invent a spending total will happily do it.
+
+// MoneyWindow — which period a money question asks about.
+type MoneyWindow int
+
+const (
+	MoneyNone MoneyWindow = iota
+	MoneyToday
+	MoneyMonth
+)
+
+// moneyNouns — the words that make a question be about his money.
+var moneyNouns = []string{
+	"потратил", "потратила", "тратил", "траты", "трат", "расходы", "расходов",
+	"заработал", "потрачено", "денег", "spend", "spent", "expenses",
+}
+
+// ParseMoneyQuery reports whether an utterance asks about spending or income,
+// and over which window. Defaults to the month: "сколько я потратил?" without a
+// period is the month-to-date question, which is the one worth answering.
+//
+// Narrow on purpose. A money noun alone is not enough — "я потратил весь день
+// на это" is him talking about his day, so an amount word or an explicit
+// question word has to be there too.
+func ParseMoneyQuery(text string) (MoneyWindow, bool) {
+	toks := planTokens(text)
+	if len(toks) == 0 {
+		return MoneyNone, false
+	}
+	hasNoun := false
+	for _, t := range toks {
+		for _, n := range moneyNouns {
+			if t == n {
+				hasNoun = true
+			}
+		}
+	}
+	if !hasNoun {
+		return MoneyNone, false
+	}
+	// "весь день", "время", "силы" — spending that is not money.
+	for _, t := range toks {
+		switch t {
+		case "день", "дня", "время", "времени", "силы", "сил", "нервы":
+			return MoneyNone, false
+		}
+	}
+	asking := hasTok(toks, "сколько") || hasTok(toks, "какие") || hasTok(toks, "покажи") ||
+		hasTok(toks, "how") || hasTok(toks, "much") || hasTok(toks, "my") ||
+		hasTok(toks, "мои") || hasTok(toks, "траты") || hasTok(toks, "расходы")
+	if !asking {
+		return MoneyNone, false
+	}
+	lower := strings.ToLower(text)
+	switch {
+	case hasTok(toks, "сегодня") || strings.Contains(lower, "today"):
+		return MoneyToday, true
+	case hasTok(toks, "месяц") || hasTok(toks, "месяце") || strings.Contains(lower, "month"):
+		return MoneyMonth, true
+	}
+	return MoneyMonth, true
+}

internal/router/money_test.go

@@ -0,0 +1,31 @@
+package router
+
+import "testing"
+
+func TestParseMoneyQuery(t *testing.T) {
+	cases := []struct {
+		in     string
+		window MoneyWindow
+		ok     bool
+	}{
+		{"сколько я потратил сегодня?", MoneyToday, true},
+		{"сколько я потратил в этом месяце?", MoneyMonth, true},
+		{"сколько я потратил?", MoneyMonth, true}, // month-to-date by default
+		{"покажи мои траты", MoneyMonth, true},
+		{"какие у меня расходы за месяц", MoneyMonth, true},
+		{"how much did I spend today", MoneyToday, true},
+		{"сколько я заработал в этом месяце", MoneyMonth, true},
+		// Not about money.
+		{"я потратил весь день на это", MoneyNone, false},
+		{"потратил много сил", MoneyNone, false},
+		{"какая погода?", MoneyNone, false},
+		{"я купил молоко", MoneyNone, false},
+		{"", MoneyNone, false},
+	}
+	for _, c := range cases {
+		w, ok := ParseMoneyQuery(c.in)
+		if ok != c.ok || w != c.window {
+			t.Errorf("ParseMoneyQuery(%q) = (%v, %v), want (%v, %v)", c.in, w, ok, c.window, c.ok)
+		}
+	}
+}

internal/router/task.go

@@ -38,11 +38,33 @@ var taskCapturePrefixes = []string{
 	"new task",
 }
 
+// TaskCapture — a parsed capture: the task itself, plus the importance he
+// stated out loud if he stated one (Vikunja #129). Weight 0 means he said
+// nothing about importance, which the ranker treats as exactly that — no
+// urgency is inferred from the wording.
+type TaskCapture struct {
+	Text   string
+	Weight int
+}
+
+// urgencyMarkers — the words that set a weight, strongest first. Only these
+// two rungs: "срочно" is a deadline he has not named, "важно" is a preference,
+// and a third shade of urgent would be a distinction he never makes out loud.
+var urgencyMarkers = []struct {
+	word   string
+	weight int
+}{
+	{"срочно", 3},
+	{"urgent", 3},
+	{"важно", 2},
+	{"important", 2},
+}
+
 // ParseTaskCapture reports whether an utterance explicitly files a task, and
 // returns the task text with the marker stripped. A marker with nothing after it
 // is not a capture (there is no task in "добавь в задачи") — the caller falls
 // through to whatever it would otherwise have done with the turn.
-func ParseTaskCapture(text string) (string, bool) {
+func ParseTaskCapture(text string) (TaskCapture, bool) {
 	trimmed := strings.TrimSpace(text)
 	lower := strings.ToLower(trimmed)
 	best := ""
@@ -52,7 +74,7 @@ func ParseTaskCapture(text string) (string, bool) {
 		}
 	}
 	if best == "" {
-		return "", false
+		return TaskCapture{}, false
 	}
 	// Cut on the rune length of the matched prefix. ToLower does not change the
 	// byte length of Russian or English letters, so the index carries over.
@@ -60,10 +82,35 @@ func ParseTaskCapture(text string) (string, bool) {
 	rest = strings.TrimLeft(rest, ":—- ")
 	rest = strings.TrimSpace(rest)
 	rest = strings.TrimRight(rest, ".!")
+	rest, weight := stripUrgency(rest)
 	if rest == "" {
-		return "", false
+		return TaskCapture{}, false
 	}
-	return rest, true
+	return TaskCapture{Text: rest, Weight: weight}, true
+}
+
+// stripUrgency pulls a leading or trailing urgency word out of the task text
+// and returns the weight it implies. Only at the edges: "срочно оплатить
+// интернет" and "оплатить интернет срочно" are the same instruction, while
+// "позвонить в срочную помощь" is a task whose text happens to contain the
+// stem, and cutting a word out of the middle of it would mangle the task.
+//
+// The word is removed from the text, because the list should read "оплатить
+// интернет (важно)" and not "важно оплатить интернет (важно)".
+func stripUrgency(text string) (string, int) {
+	for _, m := range urgencyMarkers {
+		lower := strings.ToLower(text)
+		switch {
+		case strings.HasPrefix(lower, m.word+" "):
+			return strings.TrimSpace(text[len(m.word):]), m.weight
+		case strings.HasSuffix(lower, " "+m.word):
+			return strings.TrimSpace(text[:len(text)-len(m.word)]), m.weight
+		case lower == m.word:
+			// Nothing but the marker — no task in it.
+			return "", 0
+		}
+	}
+	return text, 0
 }
 
 // taskListWords — the nouns that make a question be about the task list.

internal/router/task_test.go

@@ -4,28 +4,36 @@ import "testing"
 
 func TestParseTaskCapture(t *testing.T) {
 	cases := []struct {
-		in   string
-		text string
-		ok   bool
+		in     string
+		text   string
+		weight int
+		ok     bool
 	}{
-		{"добавь в задачи купить молоко", "купить молоко", true},
-		{"Добавь в список дел: позвонить в банк", "позвонить в банк", true},
-		{"запиши задачу починить кран.", "починить кран", true},
-		{"новая задача — оплатить интернет", "оплатить интернет", true},
-		{"add a task buy milk", "buy milk", true},
+		{"добавь в задачи купить молоко", "купить молоко", 0, true},
+		{"Добавь в список дел: позвонить в банк", "позвонить в банк", 0, true},
+		{"запиши задачу починить кран.", "починить кран", 0, true},
+		{"новая задача — оплатить интернет", "оплатить интернет", 0, true},
+		{"add a task buy milk", "buy milk", 0, true},
+		// Urgency he stated out loud, leading or trailing, stripped from the text.
+		{"добавь в задачи срочно оплатить интернет", "оплатить интернет", 3, true},
+		{"добавь в задачи оплатить интернет срочно", "оплатить интернет", 3, true},
+		{"новая задача важно позвонить маме", "позвонить маме", 2, true},
+		// The stem inside the task text is part of the task, not a marker.
+		{"добавь в задачи позвонить в срочную помощь", "позвонить в срочную помощь", 0, true},
 		// A marker with nothing after it files nothing.
-		{"добавь в задачи", "", false},
-		{"новая задача", "", false},
+		{"добавь в задачи", "", 0, false},
+		{"новая задача", "", 0, false},
+		{"добавь в задачи срочно", "", 0, false},
 		// Not a capture: he is talking, not filing.
-		{"надо бы поспать", "", false},
-		{"я не добавил молоко в список", "", false},
-		{"какие у меня задачи?", "", false},
-		{"", "", false},
+		{"надо бы поспать", "", 0, false},
+		{"я не добавил молоко в список", "", 0, false},
+		{"какие у меня задачи?", "", 0, false},
+		{"", "", 0, false},
 	}
 	for _, c := range cases {
-		text, ok := ParseTaskCapture(c.in)
-		if ok != c.ok || text != c.text {
-			t.Errorf("ParseTaskCapture(%q) = (%q, %v), want (%q, %v)", c.in, text, ok, c.text, c.ok)
+		got, ok := ParseTaskCapture(c.in)
+		if ok != c.ok || got.Text != c.text || got.Weight != c.weight {
+			t.Errorf("ParseTaskCapture(%q) = (%+v, %v), want (%q, w=%d, %v)", c.in, got, ok, c.text, c.weight, c.ok)
 		}
 	}
 }

internal/router/url.go

@@ -0,0 +1,39 @@
+package router
+
+import (
+	"regexp"
+	"strings"
+)
+
+// Finding a URL in an utterance (Vikunja #259).
+//
+// This is deliberately strict: a scheme is required. "посмотри на example.org"
+// is not treated as a fetch request, because a bare dotted word is also how
+// people say file names, versions and Russian abbreviations, and the cost of a
+// false positive here is an outbound request nobody asked for.
+//
+// Note where this runs: an utterance from STT. Whisper will mangle a spoken URL,
+// which is fine — the URL that survives is one he pasted into the web chat, and
+// a mangled one simply fails to match.
+var urlRE = regexp.MustCompile(`(?i)\bhttps?://[^\s<>"']+`)
+
+// FirstURL returns the first http(s) URL in text.
+//
+// Trailing punctuation is trimmed: he ends sentences, and "…/page." is not a
+// path component. A closing bracket is only trimmed when it has no opener,
+// because a wikipedia URL legitimately ends in one.
+func FirstURL(text string) (string, bool) {
+	m := urlRE.FindString(text)
+	if m == "" {
+		return "", false
+	}
+	m = strings.TrimRight(m, ".,;:!?…")
+	if strings.HasSuffix(m, ")") && strings.Count(m, "(") == 0 {
+		m = strings.TrimSuffix(m, ")")
+	}
+	// A scheme with nothing after it is not a URL.
+	if rest := strings.SplitN(m, "//", 2); len(rest) < 2 || rest[1] == "" {
+		return "", false
+	}
+	return m, true
+}

internal/router/url_test.go

@@ -0,0 +1,34 @@
+package router
+
+import "testing"
+
+func TestFirstURL(t *testing.T) {
+	cases := []struct {
+		text string
+		want string
+	}{
+		{"посмотри https://example.org/page — что там?", "https://example.org/page"},
+		{"почитай http://example.org/a/b?x=1 и скажи", "http://example.org/a/b?x=1"},
+		{"вот ссылка: https://example.org/page.", "https://example.org/page"},
+		{"https://ru.wikipedia.org/wiki/Небо_(значения)", "https://ru.wikipedia.org/wiki/Небо_(значения)"},
+		// No scheme ⇒ no fetch. A bare dotted word is not an instruction to
+		// reach out to the network.
+		{"посмотри на example.org", ""},
+		{"открой файл config.json", ""},
+		{"что нового?", ""},
+		{"https://", ""},
+		{"", ""},
+	}
+	for _, c := range cases {
+		got, ok := FirstURL(c.text)
+		if c.want == "" {
+			if ok {
+				t.Errorf("FirstURL(%q) = %q, want no match", c.text, got)
+			}
+			continue
+		}
+		if !ok || got != c.want {
+			t.Errorf("FirstURL(%q) = %q, %v; want %q", c.text, got, ok, c.want)
+		}
+	}
+}

internal/rss/feed.go

@@ -0,0 +1,179 @@
+// Package rss reads RSS 2.0 and Atom feeds, and does nothing else with them.
+//
+// Parsing and polling are split from delivery on purpose: a feed is a source
+// Maven can be ASKED about, not a thing that speaks. Nothing in this package
+// dispatches, nudges or notifies — the poller writes notes, and the answer path
+// reads them when he asks "что нового в лентах?". "Not a nag" is the oldest
+// constraint in the spec, and a news feed is the single most tempting way to
+// break it.
+//
+// Stdlib only (encoding/xml). Feeds are XML from strangers, so the parser takes
+// what it recognises and ignores the rest rather than failing a whole feed over
+// one malformed item.
+package rss
+
+import (
+	"encoding/xml"
+	"fmt"
+	"html"
+	"io"
+	"regexp"
+	"strings"
+	"time"
+)
+
+// Item is one feed entry, normalised across RSS and Atom.
+type Item struct {
+	Title     string
+	Link      string
+	Summary   string    // plain text, tags stripped, entities decoded
+	Published time.Time // zero when the feed did not say
+	ID        string    // guid / atom id, falling back to the link
+}
+
+// Feed is a parsed document.
+type Feed struct {
+	Title string
+	Items []Item
+}
+
+// feedDoc covers both dialects in one struct. RSS puts items under
+// channel>item, Atom puts entries at the top level, and the field names barely
+// overlap — so both sets are declared and whichever the document filled in wins.
+type feedDoc struct {
+	ChannelTitle string `xml:"channel>title"`
+	AtomTitle    string `xml:"title"`
+
+	Items []struct {
+		Title       string `xml:"title"`
+		Link        string `xml:"link"`
+		Description string `xml:"description"`
+		Encoded     string `xml:"encoded"` // content:encoded
+		GUID        string `xml:"guid"`
+		PubDate     string `xml:"pubDate"`
+		Date        string `xml:"date"` // dc:date
+	} `xml:"channel>item"`
+
+	Entries []struct {
+		Title string `xml:"title"`
+		Links []struct {
+			Href string `xml:"href,attr"`
+			Rel  string `xml:"rel,attr"`
+		} `xml:"link"`
+		Summary   string `xml:"summary"`
+		Content   string `xml:"content"`
+		ID        string `xml:"id"`
+		Updated   string `xml:"updated"`
+		Published string `xml:"published"`
+	} `xml:"entry"`
+}
+
+// Parse reads a feed document.
+func Parse(r io.Reader) (Feed, error) {
+	var doc feedDoc
+	dec := xml.NewDecoder(r)
+	// Feeds in the wild declare windows-1251 and worse. We only ever read
+	// UTF-8; a charset we cannot decode is a feed we do not read, which is
+	// better than mojibake in his notes.
+	dec.Strict = false
+	if err := dec.Decode(&doc); err != nil {
+		return Feed{}, fmt.Errorf("rss: bad xml: %w", err)
+	}
+
+	f := Feed{Title: strings.TrimSpace(doc.ChannelTitle)}
+	if f.Title == "" {
+		f.Title = strings.TrimSpace(doc.AtomTitle)
+	}
+	for _, it := range doc.Items {
+		item := Item{
+			Title:     PlainText(it.Title),
+			Link:      strings.TrimSpace(it.Link),
+			Summary:   PlainText(firstNonEmpty(it.Description, it.Encoded)),
+			Published: parseTime(firstNonEmpty(it.PubDate, it.Date)),
+			ID:        strings.TrimSpace(firstNonEmpty(it.GUID, it.Link)),
+		}
+		if item.Title != "" || item.Link != "" {
+			f.Items = append(f.Items, item)
+		}
+	}
+	for _, e := range doc.Entries {
+		link := ""
+		for _, l := range e.Links {
+			if l.Rel == "" || l.Rel == "alternate" {
+				link = strings.TrimSpace(l.Href)
+				break
+			}
+		}
+		if link == "" && len(e.Links) > 0 {
+			link = strings.TrimSpace(e.Links[0].Href)
+		}
+		item := Item{
+			Title:     PlainText(e.Title),
+			Link:      link,
+			Summary:   PlainText(firstNonEmpty(e.Summary, e.Content)),
+			Published: parseTime(firstNonEmpty(e.Published, e.Updated)),
+			ID:        strings.TrimSpace(firstNonEmpty(e.ID, link)),
+		}
+		if item.Title != "" || item.Link != "" {
+			f.Items = append(f.Items, item)
+		}
+	}
+	return f, nil
+}
+
+func firstNonEmpty(vals ...string) string {
+	for _, v := range vals {
+		if strings.TrimSpace(v) != "" {
+			return v
+		}
+	}
+	return ""
+}
+
+// timeLayouts — RFC1123/822 for RSS, RFC3339 for Atom, plus the near-misses
+// real feeds ship (no seconds, numeric zone where a name is expected).
+var timeLayouts = []string{
+	time.RFC1123Z,
+	time.RFC1123,
+	time.RFC822Z,
+	time.RFC822,
+	time.RFC3339,
+	"2006-01-02T15:04:05Z0700",
+	"2006-01-02 15:04:05",
+	"2006-01-02",
+	"Mon, 02 Jan 2006 15:04:05 -0700",
+	"Mon, 2 Jan 2006 15:04:05 -0700",
+	"Mon, 2 Jan 2006 15:04:05 MST",
+}
+
+// parseTime returns the zero time on anything it cannot read. An undated item
+// is still an item; the poller dedupes by ID, so a missing date costs nothing.
+func parseTime(s string) time.Time {
+	s = strings.TrimSpace(s)
+	if s == "" {
+		return time.Time{}
+	}
+	for _, l := range timeLayouts {
+		if t, err := time.Parse(l, s); err == nil {
+			return t.UTC()
+		}
+	}
+	return time.Time{}
+}
+
+var (
+	// RE2 has no backreferences, so the two tags are spelled out rather than
+	// captured and matched against themselves.
+	scriptRE = regexp.MustCompile(`(?is)<script\b[^>]*>.*?</script>|<style\b[^>]*>.*?</style>`)
+	tagRE    = regexp.MustCompile(`(?s)<[^>]*>`)
+)
+
+// PlainText strips markup and decodes entities — feed summaries are HTML, and
+// what reaches a note (and possibly the TTS) must be text. Exported because the
+// crawler's extractor needs exactly this on a bigger input.
+func PlainText(s string) string {
+	s = scriptRE.ReplaceAllString(s, " ")
+	s = tagRE.ReplaceAllString(s, " ")
+	s = html.UnescapeString(s)
+	return strings.TrimSpace(strings.Join(strings.Fields(s), " "))
+}

internal/rss/feed_test.go

@@ -0,0 +1,112 @@
+package rss
+
+import (
+	"strings"
+	"testing"
+	"time"
+)
+
+const rss2 = `<?xml version="1.0"?>
+<rss version="2.0">
+ <channel>
+  <title>Хабр</title>
+  <item>
+   <title>Новая уязвимость в ядре</title>
+   <link>https://example.org/a</link>
+   <description>&lt;p&gt;Патч уже &lt;b&gt;вышел&lt;/b&gt;.&lt;/p&gt;</description>
+   <guid>tag:example.org,a</guid>
+   <pubDate>Mon, 28 Jul 2026 10:00:00 +0000</pubDate>
+  </item>
+  <item>
+   <title>Без даты</title>
+   <link>https://example.org/b</link>
+  </item>
+ </channel>
+</rss>`
+
+const atom = `<?xml version="1.0" encoding="utf-8"?>
+<feed xmlns="http://www.w3.org/2005/Atom">
+ <title>Example Atom</title>
+ <entry>
+  <title>Release 2.0</title>
+  <link rel="alternate" href="https://example.com/rel"/>
+  <link rel="edit" href="https://example.com/edit"/>
+  <id>urn:uuid:1</id>
+  <updated>2026-07-30T12:30:00Z</updated>
+  <summary>Ships &amp; works</summary>
+ </entry>
+</feed>`
+
+func TestParseRSS2(t *testing.T) {
+	f, err := Parse(strings.NewReader(rss2))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if f.Title != "Хабр" {
+		t.Fatalf("title = %q", f.Title)
+	}
+	if len(f.Items) != 2 {
+		t.Fatalf("items = %d, want 2", len(f.Items))
+	}
+	it := f.Items[0]
+	if it.Title != "Новая уязвимость в ядре" {
+		t.Errorf("title = %q", it.Title)
+	}
+	if it.Summary != "Патч уже вышел ." && it.Summary != "Патч уже вышел." {
+		t.Errorf("summary = %q — tags must be stripped and entities decoded", it.Summary)
+	}
+	if it.ID != "tag:example.org,a" {
+		t.Errorf("id = %q", it.ID)
+	}
+	if want := time.Date(2026, 7, 28, 10, 0, 0, 0, time.UTC); !it.Published.Equal(want) {
+		t.Errorf("published = %v, want %v", it.Published, want)
+	}
+	if !f.Items[1].Published.IsZero() {
+		t.Errorf("undated item got a date: %v", f.Items[1].Published)
+	}
+	if f.Items[1].ID != "https://example.org/b" {
+		t.Errorf("id falls back to the link, got %q", f.Items[1].ID)
+	}
+}
+
+func TestParseAtom(t *testing.T) {
+	f, err := Parse(strings.NewReader(atom))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if f.Title != "Example Atom" || len(f.Items) != 1 {
+		t.Fatalf("feed = %+v", f)
+	}
+	it := f.Items[0]
+	if it.Link != "https://example.com/rel" {
+		t.Errorf("link = %q, want the alternate link", it.Link)
+	}
+	if it.Summary != "Ships & works" {
+		t.Errorf("summary = %q", it.Summary)
+	}
+	if want := time.Date(2026, 7, 30, 12, 30, 0, 0, time.UTC); !it.Published.Equal(want) {
+		t.Errorf("published = %v, want %v", it.Published, want)
+	}
+}
+
+func TestParseGarbage(t *testing.T) {
+	if _, err := Parse(strings.NewReader("<html><body>not a feed")); err == nil {
+		t.Fatal("want an error on a non-feed document")
+	}
+	// A feed with an item that has neither title nor link contributes nothing
+	// rather than an empty note.
+	f, err := Parse(strings.NewReader(`<rss><channel><item><description>x</description></item></channel></rss>`))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(f.Items) != 0 {
+		t.Fatalf("items = %d, want 0", len(f.Items))
+	}
+}
+
+func TestPlainTextDropsScript(t *testing.T) {
+	got := PlainText(`<p>hi</p><script>alert("x")</script><style>b{}</style> there`)
+	if got != "hi there" {
+		t.Fatalf("got %q", got)
+	}
+}

internal/rss/poller.go

@@ -0,0 +1,324 @@
+package rss
+
+import (
+	"context"
+	"fmt"
+	"log"
+	"strings"
+	"time"
+)
+
+// FeedConfig — one feed to read. A feed with no Name or no URL is ignored.
+type FeedConfig struct {
+	Name     string        // short id; the note source is "rss:<Name>"
+	URL      string        // http(s) only, enforced by the fetcher
+	Category string        // free text ("технологии"), used to answer "что по X?"
+	Interval time.Duration // 0 ⇒ the poller's default
+	Include  []string      // when non-empty, keep only items matching one of these
+	Exclude  []string      // drop items matching any of these, even if included
+}
+
+// Fetcher is the guarded HTTP door (internal/webfetch). An interface so the
+// poller is testable without a network and so it CANNOT fetch by any other
+// means: no http.Client is constructed in this package.
+type Fetcher interface {
+	Get(ctx context.Context, url string) (*Body, error)
+}
+
+// Body is the minimum the poller needs from a response.
+type Body struct{ Bytes []byte }
+
+// Notes is core's note-writing half. Same shape as ipc.CoreAPI's method, so the
+// daemon passes its API straight in.
+type Notes interface {
+	WriteNote(ctx context.Context, ts time.Time, text string, embedding []float32, source string) (int64, error)
+}
+
+// Marks remembers how far a feed was read. Durable, because the alternative is
+// re-writing yesterday's headlines as fresh notes after every restart. The
+// daemon backs this with config facts (key "rss:latest:<feed>").
+type Marks interface {
+	LastMark(ctx context.Context, feed string) (time.Time, error)
+	SetMark(ctx context.Context, feed string, at time.Time) error
+}
+
+// Embedder embeds a note on its way into the store so recall can find it. nil ⇒
+// notes are written without a vector (still readable by the recent-notes path).
+type Embedder interface {
+	Embed(ctx context.Context, text string) ([]float32, error)
+}
+
+// Ranker is the relevance seam. The plan called for scoring each item against
+// an interest profile built from his notes; that profile does not exist yet, and
+// a threshold over an embedder with no profile to compare to is a random filter
+// with a confident name. So the seam is here, nil in the daemon, and the filter
+// that actually runs is the per-feed keyword one — a rule he can read and
+// predict. When there IS a profile, implement this and pass it.
+//
+// Note what a Ranker must NOT be: anything that sends his notes outward. The
+// scoring happens locally against a local embedder; the feed item is the input,
+// his memory is never the payload.
+type Ranker interface {
+	Relevant(ctx context.Context, text string) (bool, error)
+}
+
+// Config — poller-wide settings.
+type Config struct {
+	DefaultInterval time.Duration // 0 ⇒ DefaultPollInterval
+	MaxItems        int           // most notes written per feed per poll; 0 ⇒ DefaultMaxItems
+	MaxAge          time.Duration // ignore items older than this on a cold start; 0 ⇒ DefaultMaxAge
+}
+
+// Defaults chosen to be quiet: a feed read every half hour, at most a handful of
+// items kept, and a cold start that does not import a month of history.
+const (
+	DefaultPollInterval = 30 * time.Minute
+	DefaultMaxItems     = 5
+	DefaultMaxAge       = 24 * time.Hour
+)
+
+// Poller reads feeds on a schedule and writes what survives filtering as notes.
+type Poller struct {
+	feeds   []FeedConfig
+	fetch   Fetcher
+	notes   Notes
+	marks   Marks
+	embed   Embedder
+	ranker  Ranker
+	cfg     Config
+	nextDue map[string]time.Time
+	seen    map[string]map[string]bool // feed → item ID, for items with no date
+}
+
+// NewPoller wires a poller. Returns nil when there is nothing to poll — a
+// capability is off unless configured, and callers check for nil.
+func NewPoller(feeds []FeedConfig, fetch Fetcher, notes Notes, marks Marks, embed Embedder, ranker Ranker, cfg Config) *Poller {
+	var valid []FeedConfig
+	for _, f := range feeds {
+		if strings.TrimSpace(f.Name) == "" || strings.TrimSpace(f.URL) == "" {
+			log.Printf("rss: skipping a feed with no name or no url")
+			continue
+		}
+		valid = append(valid, f)
+	}
+	if len(valid) == 0 || fetch == nil || notes == nil {
+		return nil
+	}
+	if cfg.DefaultInterval <= 0 {
+		cfg.DefaultInterval = DefaultPollInterval
+	}
+	if cfg.MaxItems <= 0 {
+		cfg.MaxItems = DefaultMaxItems
+	}
+	if cfg.MaxAge <= 0 {
+		cfg.MaxAge = DefaultMaxAge
+	}
+	return &Poller{
+		feeds: valid, fetch: fetch, notes: notes, marks: marks,
+		embed: embed, ranker: ranker, cfg: cfg,
+		nextDue: map[string]time.Time{},
+		seen:    map[string]map[string]bool{},
+	}
+}
+
+// Feeds returns the configured feeds (the answer path lists categories).
+func (p *Poller) Feeds() []FeedConfig { return p.feeds }
+
+// PollDue reads every feed whose interval has elapsed and returns how many
+// notes were written. Errors are logged per feed, never returned: one dead feed
+// must not stop the others, and there is nobody waiting on this.
+func (p *Poller) PollDue(ctx context.Context, now time.Time) int {
+	written := 0
+	for _, f := range p.feeds {
+		if due, ok := p.nextDue[f.Name]; ok && now.Before(due) {
+			continue
+		}
+		interval := f.Interval
+		if interval <= 0 {
+			interval = p.cfg.DefaultInterval
+		}
+		p.nextDue[f.Name] = now.Add(interval)
+		n, err := p.PollFeed(ctx, f, now)
+		if err != nil {
+			// The URL is configured by him and not a secret, so it is loggable;
+			// item titles are not logged, only counts.
+			log.Printf("rss: feed %s: %v", f.Name, err)
+			continue
+		}
+		if n > 0 {
+			log.Printf("rss: feed %s: %d new item(s) noted", f.Name, n)
+		}
+		written += n
+	}
+	return written
+}
+
+// PollFeed reads one feed now, regardless of its schedule.
+func (p *Poller) PollFeed(ctx context.Context, f FeedConfig, now time.Time) (int, error) {
+	body, err := p.fetch.Get(ctx, f.URL)
+	if err != nil {
+		return 0, err
+	}
+	feed, err := Parse(strings.NewReader(string(body.Bytes)))
+	if err != nil {
+		return 0, err
+	}
+
+	mark := p.mark(ctx, f.Name, now)
+	newest := mark
+	written := 0
+	for _, it := range feed.Items {
+		if written >= p.cfg.MaxItems {
+			break
+		}
+		if !p.fresh(f, it, mark, now) {
+			continue
+		}
+		if !Matches(f, it) {
+			continue
+		}
+		if p.ranker != nil {
+			ok, err := p.ranker.Relevant(ctx, it.Title+" "+it.Summary)
+			if err != nil {
+				log.Printf("rss: feed %s: relevance: %v", f.Name, err)
+			} else if !ok {
+				continue
+			}
+		}
+		if err := p.write(ctx, f, it, now); err != nil {
+			return written, err
+		}
+		written++
+		if it.Published.After(newest) {
+			newest = it.Published
+		}
+	}
+	if p.marks != nil && newest.After(mark) {
+		if err := p.marks.SetMark(ctx, f.Name, newest); err != nil {
+			log.Printf("rss: feed %s: save mark: %v", f.Name, err)
+		}
+	}
+	return written, nil
+}
+
+// mark — how far this feed was read. A feed with no mark starts MaxAge ago, so
+// a first poll takes today's headlines instead of the whole archive.
+func (p *Poller) mark(ctx context.Context, feed string, now time.Time) time.Time {
+	cold := now.Add(-p.cfg.MaxAge)
+	if p.marks == nil {
+		return cold
+	}
+	at, err := p.marks.LastMark(ctx, feed)
+	if err != nil || at.IsZero() {
+		return cold
+	}
+	return at
+}
+
+// fresh — two dedup rules, because feeds are inconsistent about dates. A dated
+// item must be newer than the mark; an undated one is kept once per process by
+// ID. Both are needed: dates alone re-import undated feeds forever, IDs alone
+// lose their memory on restart.
+func (p *Poller) fresh(f FeedConfig, it Item, mark, now time.Time) bool {
+	if !it.Published.IsZero() {
+		if !it.Published.After(mark) {
+			return false
+		}
+		// A feed that dates its items in the future (or a clock skew) must not
+		// win the mark and mute everything after it.
+		return !it.Published.After(now.Add(time.Hour))
+	}
+	id := it.ID
+	if id == "" {
+		id = it.Title
+	}
+	if p.seen[f.Name] == nil {
+		p.seen[f.Name] = map[string]bool{}
+	}
+	if p.seen[f.Name][id] {
+		return false
+	}
+	p.seen[f.Name][id] = true
+	return true
+}
+
+// write stores one item as a note. Source "rss:<feed>" is what the answer path
+// filters on, and what makes a feed note distinguishable from something he said.
+func (p *Poller) write(ctx context.Context, f FeedConfig, it Item, now time.Time) error {
+	text := NoteText(f, it)
+	var vec []float32
+	if p.embed != nil {
+		v, err := p.embed.Embed(ctx, text)
+		if err != nil {
+			log.Printf("rss: feed %s: embed: %v", f.Name, err)
+		} else {
+			vec = v
+		}
+	}
+	ts := it.Published
+	if ts.IsZero() {
+		ts = now
+	}
+	if _, err := p.notes.WriteNote(ctx, ts, text, vec, SourceFor(f.Name)); err != nil {
+		return fmt.Errorf("write note: %w", err)
+	}
+	return nil
+}
+
+// SourceFor is the note source for a feed.
+func SourceFor(feed string) string { return "rss:" + feed }
+
+// SourcePrefix — what the answer path matches to find feed notes.
+const SourcePrefix = "rss:"
+
+// NoteText renders an item as the note body. The category is included because
+// "что нового по технологиям?" is answered by reading notes, and a note has to
+// carry enough to be recognised as belonging to that category.
+func NoteText(f FeedConfig, it Item) string {
+	var b strings.Builder
+	b.WriteString(it.Title)
+	if f.Category != "" {
+		fmt.Fprintf(&b, " [%s]", f.Category)
+	}
+	if it.Summary != "" {
+		b.WriteString("\n")
+		b.WriteString(trimRunes(it.Summary, 500))
+	}
+	if it.Link != "" {
+		b.WriteString("\n")
+		b.WriteString(it.Link)
+	}
+	return b.String()
+}
+
+// trimRunes cuts on a rune boundary — a note is Russian as often as English and
+// half a cyrillic letter is a broken note.
+func trimRunes(s string, max int) string {
+	r := []rune(s)
+	if len(r) <= max {
+		return s
+	}
+	return strings.TrimSpace(string(r[:max])) + "…"
+}
+
+// Matches applies the per-feed keyword filter: keep when Include is empty or one
+// include matches, drop when any exclude matches. Case-insensitive substring,
+// which for Russian is the honest choice — no stemmer here, so "выборы" does not
+// match "выборах", and a filter he writes is a filter he can predict.
+func Matches(f FeedConfig, it Item) bool {
+	hay := strings.ToLower(it.Title + " " + it.Summary)
+	for _, x := range f.Exclude {
+		if x = strings.ToLower(strings.TrimSpace(x)); x != "" && strings.Contains(hay, x) {
+			return false
+		}
+	}
+	if len(f.Include) == 0 {
+		return true
+	}
+	for _, in := range f.Include {
+		if in = strings.ToLower(strings.TrimSpace(in)); in != "" && strings.Contains(hay, in) {
+			return true
+		}
+	}
+	return false
+}

internal/rss/poller_test.go

@@ -0,0 +1,210 @@
+package rss
+
+import (
+	"context"
+	"errors"
+	"strings"
+	"testing"
+	"time"
+)
+
+type fakeFetch struct {
+	body  string
+	err   error
+	calls int
+	urls  []string
+}
+
+func (f *fakeFetch) Get(_ context.Context, url string) (*Body, error) {
+	f.calls++
+	f.urls = append(f.urls, url)
+	if f.err != nil {
+		return nil, f.err
+	}
+	return &Body{Bytes: []byte(f.body)}, nil
+}
+
+type writtenNote struct {
+	ts     time.Time
+	text   string
+	source string
+	vec    []float32
+}
+
+type fakeNotes struct{ notes []writtenNote }
+
+func (n *fakeNotes) WriteNote(_ context.Context, ts time.Time, text string, vec []float32, source string) (int64, error) {
+	n.notes = append(n.notes, writtenNote{ts, text, source, vec})
+	return int64(len(n.notes)), nil
+}
+
+type fakeMarks struct{ m map[string]time.Time }
+
+func newMarks() *fakeMarks { return &fakeMarks{m: map[string]time.Time{}} }
+func (f *fakeMarks) LastMark(_ context.Context, feed string) (time.Time, error) {
+	return f.m[feed], nil
+}
+func (f *fakeMarks) SetMark(_ context.Context, feed string, at time.Time) error {
+	f.m[feed] = at
+	return nil
+}
+
+var now = time.Date(2026, 7, 28, 12, 0, 0, 0, time.UTC)
+
+func TestPollWritesNotesWithSource(t *testing.T) {
+	fetch := &fakeFetch{body: rss2}
+	notes := &fakeNotes{}
+	marks := newMarks()
+	p := NewPoller([]FeedConfig{{Name: "habr", URL: "https://example.org/rss", Category: "технологии"}},
+		fetch, notes, marks, nil, nil, Config{})
+	if p == nil {
+		t.Fatal("NewPoller returned nil for a configured feed")
+	}
+	n := p.PollDue(context.Background(), now)
+	if n != 2 || len(notes.notes) != 2 {
+		t.Fatalf("wrote %d notes (returned %d), want 2", len(notes.notes), n)
+	}
+	if notes.notes[0].source != "rss:habr" {
+		t.Errorf("source = %q, want rss:habr", notes.notes[0].source)
+	}
+	if !strings.Contains(notes.notes[0].text, "технологии") {
+		t.Errorf("note does not carry its category: %q", notes.notes[0].text)
+	}
+	if !strings.Contains(notes.notes[0].text, "https://example.org/a") {
+		t.Errorf("note does not carry its link: %q", notes.notes[0].text)
+	}
+	// The undated item is stamped with now, the dated one with its own date.
+	if !notes.notes[1].ts.Equal(now) {
+		t.Errorf("undated item ts = %v, want now", notes.notes[1].ts)
+	}
+}
+
+// The whole point of a mark: polling twice must not re-note the same headlines.
+func TestSecondPollIsQuiet(t *testing.T) {
+	fetch := &fakeFetch{body: rss2}
+	notes := &fakeNotes{}
+	p := NewPoller([]FeedConfig{{Name: "habr", URL: "u", Interval: time.Minute}}, fetch, notes, newMarks(), nil, nil, Config{})
+	p.PollDue(context.Background(), now)
+	before := len(notes.notes)
+	p.PollDue(context.Background(), now.Add(2*time.Minute))
+	if len(notes.notes) != before {
+		t.Fatalf("second poll wrote %d extra notes", len(notes.notes)-before)
+	}
+}
+
+// A mark that survives a restart is the durable half; simulate one by building a
+// fresh poller over the same marks.
+func TestMarkSurvivesRestart(t *testing.T) {
+	marks := newMarks()
+	fetch := &fakeFetch{body: rss2}
+	notes := &fakeNotes{}
+	feeds := []FeedConfig{{Name: "habr", URL: "u"}}
+	NewPoller(feeds, fetch, notes, marks, nil, nil, Config{}).PollDue(context.Background(), now)
+	if len(notes.notes) != 2 {
+		t.Fatalf("first run wrote %d", len(notes.notes))
+	}
+	notes2 := &fakeNotes{}
+	NewPoller(feeds, fetch, notes2, marks, nil, nil, Config{}).PollDue(context.Background(), now.Add(time.Hour))
+	// The dated item is behind the mark. The undated one has no date to compare,
+	// so it comes back — accepted and documented in fresh(): an undated feed is
+	// deduped per process, not forever.
+	for _, n := range notes2.notes {
+		if strings.Contains(n.text, "уязвимость") {
+			t.Fatalf("dated item re-noted after restart: %q", n.text)
+		}
+	}
+}
+
+func TestIntervalIsRespected(t *testing.T) {
+	fetch := &fakeFetch{body: rss2}
+	p := NewPoller([]FeedConfig{{Name: "habr", URL: "u", Interval: time.Hour}}, fetch, &fakeNotes{}, newMarks(), nil, nil, Config{})
+	p.PollDue(context.Background(), now)
+	p.PollDue(context.Background(), now.Add(time.Minute))
+	if fetch.calls != 1 {
+		t.Fatalf("fetched %d times inside one interval, want 1", fetch.calls)
+	}
+	p.PollDue(context.Background(), now.Add(2*time.Hour))
+	if fetch.calls != 2 {
+		t.Fatalf("fetched %d times, want 2 after the interval elapsed", fetch.calls)
+	}
+}
+
+func TestColdStartIgnoresOldItems(t *testing.T) {
+	old := `<rss><channel><item><title>Старое</title><link>l</link>` +
+		`<pubDate>Mon, 01 Jun 2026 10:00:00 +0000</pubDate></item></channel></rss>`
+	notes := &fakeNotes{}
+	p := NewPoller([]FeedConfig{{Name: "f", URL: "u"}}, &fakeFetch{body: old}, notes, newMarks(), nil, nil, Config{MaxAge: 24 * time.Hour})
+	if n := p.PollDue(context.Background(), now); n != 0 {
+		t.Fatalf("cold start imported %d old items, want 0", n)
+	}
+}
+
+func TestMaxItemsCap(t *testing.T) {
+	var b strings.Builder
+	b.WriteString("<rss><channel>")
+	for i := 0; i < 10; i++ {
+		b.WriteString("<item><title>t")
+		b.WriteByte(byte('0' + i))
+		b.WriteString("</title><link>https://example.org/")
+		b.WriteByte(byte('0' + i))
+		b.WriteString("</link></item>")
+	}
+	b.WriteString("</channel></rss>")
+	notes := &fakeNotes{}
+	p := NewPoller([]FeedConfig{{Name: "f", URL: "u"}}, &fakeFetch{body: b.String()}, notes, newMarks(), nil, nil, Config{MaxItems: 3})
+	if n := p.PollDue(context.Background(), now); n != 3 {
+		t.Fatalf("wrote %d notes, want the cap of 3", n)
+	}
+}
+
+func TestKeywordFilter(t *testing.T) {
+	f := FeedConfig{Include: []string{"ядр"}, Exclude: []string{"реклама"}}
+	if !Matches(f, Item{Title: "Новое ядро"}) {
+		t.Error("include did not match")
+	}
+	if Matches(f, Item{Title: "Новое ядро", Summary: "Реклама внутри"}) {
+		t.Error("exclude must win over include")
+	}
+	if Matches(f, Item{Title: "Погода"}) {
+		t.Error("non-matching item passed the include filter")
+	}
+	if !Matches(FeedConfig{}, Item{Title: "что угодно"}) {
+		t.Error("an unfiltered feed must keep everything")
+	}
+}
+
+type fakeRanker struct{ keep bool }
+
+func (r fakeRanker) Relevant(context.Context, string) (bool, error) { return r.keep, nil }
+
+func TestRankerCanDropEverything(t *testing.T) {
+	notes := &fakeNotes{}
+	p := NewPoller([]FeedConfig{{Name: "f", URL: "u"}}, &fakeFetch{body: rss2}, notes, newMarks(), nil, fakeRanker{false}, Config{})
+	if n := p.PollDue(context.Background(), now); n != 0 {
+		t.Fatalf("ranker rejected everything but %d notes were written", n)
+	}
+}
+
+func TestFetchErrorIsSurvivable(t *testing.T) {
+	notes := &fakeNotes{}
+	p := NewPoller([]FeedConfig{
+		{Name: "dead", URL: "u1"},
+		{Name: "live", URL: "u2"},
+	}, &fakeFetch{err: errors.New("boom")}, notes, newMarks(), nil, nil, Config{})
+	if n := p.PollDue(context.Background(), now); n != 0 {
+		t.Fatalf("n = %d", n)
+	}
+	// Both feeds were attempted: one dead feed does not abort the round.
+	if p.nextDue["live"].IsZero() {
+		t.Fatal("the second feed was never attempted")
+	}
+}
+
+func TestNoFeedsMeansNoPoller(t *testing.T) {
+	if p := NewPoller(nil, &fakeFetch{}, &fakeNotes{}, nil, nil, nil, Config{}); p != nil {
+		t.Fatal("NewPoller must return nil when nothing is configured")
+	}
+	if p := NewPoller([]FeedConfig{{Name: "", URL: ""}}, &fakeFetch{}, &fakeNotes{}, nil, nil, nil, Config{}); p != nil {
+		t.Fatal("a feed with no name or url is not a configuration")
+	}
+}

internal/tasks/rank.go

@@ -0,0 +1,238 @@
+// Package tasks ranks captured work (Vikunja #129).
+//
+// The ordering is COMPUTED, not generated. Asking a 1.7B model which of his
+// tasks matters most would produce a fluent opinion about his life with no
+// basis in anything, and a confidently wrong priority is worse than no
+// priority at all — the same reasoning as the behaviour profile in
+// internal/memory, which counts instead of summarising.
+//
+// So: four signals, all of them things he told her, and a reason string naming
+// the one that decided each row. Nothing here invents urgency. A task with no
+// due date and no weight scores nothing and sits where its age puts it, which
+// is the honest answer to "which of these matters?" when he never said.
+//
+// Ranking is a READ. It sorts and renders; it never writes, schedules or
+// announces. Maven is not a nag: a task rising to the top of this list is not a
+// reason to speak, only the order she recites in when asked.
+package tasks
+
+import (
+	"fmt"
+	"sort"
+	"strings"
+	"time"
+)
+
+// Status values, mirroring internal/store so a caller can rank ipc.Task rows
+// without importing the store.
+const (
+	StatusCandidate = "candidate"
+	StatusOpen      = "open"
+)
+
+// Item — one task to rank. The subset of a task that ranking depends on;
+// callers map their own row type onto it.
+type Item struct {
+	ID      int64
+	Text    string
+	Status  string
+	Created time.Time
+	Due     *time.Time
+	Weight  int
+}
+
+// Ranked — one task with its score and the reason that decided it.
+type Ranked struct {
+	Item
+	Score float64
+	// Reason — the dominant signal, in Russian, for the page and the spoken
+	// list. Empty when nothing distinguished this task: no due date, no
+	// weight, not old. Saying "потому что" about a task he never prioritised
+	// would be making something up.
+	Reason string
+}
+
+// Scoring weights. Deliberately coarse round numbers: this is a knob, not
+// math, and the only property that has to hold is the ordering between classes
+// (overdue beats today beats this week beats undated).
+const (
+	scoreOverdue     = 100 // he already missed it
+	scoreOverduePer  = 5   // per further day late, capped
+	scoreOverdueCap  = 40
+	scoreDueToday    = 60
+	scoreDueTomorrow = 40
+	scoreDueWeek     = 20
+	scoreDueLater    = 5
+	scorePerWeight   = 15 // "срочно" / "важно" / the web form's select
+	scorePerWeekOld  = 1  // so nothing rots at the bottom forever
+	scoreAgeCap      = 10
+	// MaxWeight — the highest importance hint capture accepts. Three rungs is
+	// as many as anyone can rank by hand honestly.
+	MaxWeight = 3
+)
+
+// Rank scores every item and returns them ordered: confirmed work first, then
+// candidates, each by score descending, oldest first on a tie.
+//
+// Candidates never outrank open work, whatever their due date. A task Maven
+// derived from something she read is a suggestion until he confirms it, and
+// putting her guess above his own stated work would be reading his priorities
+// back to him wrong.
+func Rank(items []Item, now time.Time) []Ranked {
+	out := make([]Ranked, 0, len(items))
+	for _, it := range items {
+		score, reason := score(it, now)
+		out = append(out, Ranked{Item: it, Score: score, Reason: reason})
+	}
+	sort.SliceStable(out, func(i, j int) bool {
+		ci, cj := out[i].Status == StatusCandidate, out[j].Status == StatusCandidate
+		if ci != cj {
+			return !ci // open before candidate
+		}
+		if out[i].Score != out[j].Score {
+			return out[i].Score > out[j].Score
+		}
+		return out[i].Created.Before(out[j].Created) // oldest first, FIFO
+	})
+	return out
+}
+
+// score — the per-item scoring function. Returns the score and the dominant
+// reason. Deadline beats weight when both are present: a date is a fact about
+// the world, a weight is how he felt when he filed it.
+func score(it Item, now time.Time) (float64, string) {
+	var total float64
+	reason := ""
+
+	if it.Due != nil {
+		days := dayDelta(*it.Due, now)
+		switch {
+		case days < 0:
+			late := -days
+			bonus := float64(late * scoreOverduePer)
+			if bonus > scoreOverdueCap {
+				bonus = scoreOverdueCap
+			}
+			total += scoreOverdue + bonus
+			reason = "просрочено"
+			if late == 1 {
+				reason = "просрочено на день"
+			} else if late > 1 {
+				reason = fmt.Sprintf("просрочено на %d дн.", late)
+			}
+		case days == 0:
+			total += scoreDueToday
+			reason = "сегодня"
+		case days == 1:
+			total += scoreDueTomorrow
+			reason = "завтра"
+		case days <= 7:
+			total += scoreDueWeek
+			reason = fmt.Sprintf("через %d дн.", days)
+		default:
+			total += scoreDueLater
+		}
+	}
+
+	w := it.Weight
+	if w > MaxWeight {
+		w = MaxWeight
+	}
+	if w > 0 {
+		total += float64(w * scorePerWeight)
+		if reason == "" {
+			reason = "важно"
+		}
+	}
+
+	if !it.Created.IsZero() {
+		weeks := int(now.Sub(it.Created).Hours() / (24 * 7))
+		if weeks > 0 {
+			age := float64(weeks * scorePerWeekOld)
+			if age > scoreAgeCap {
+				age = scoreAgeCap
+			}
+			total += age
+			if reason == "" && weeks >= 2 {
+				reason = "давно в списке"
+			}
+		}
+	}
+	return total, reason
+}
+
+// dayDelta — calendar days from now to due, in due's own location. Whole days,
+// not hours: a task due today is due today whether it is 09:00 or 23:00, and an
+// hours-based comparison would call this evening's task "overdue" all afternoon.
+func dayDelta(due, now time.Time) int {
+	loc := due.Location()
+	d := time.Date(due.Year(), due.Month(), due.Day(), 0, 0, 0, 0, loc)
+	n := now.In(loc)
+	n = time.Date(n.Year(), n.Month(), n.Day(), 0, 0, 0, 0, loc)
+	return int(d.Sub(n).Hours() / 24)
+}
+
+// SpokenLimit — how many tasks the spoken list names before it summarises the
+// rest. A recital of twenty items is noise; five is a list he can hold.
+const SpokenLimit = 5
+
+// FormatRU renders a ranked list the way Maven says it. Confirmed work first,
+// with the reason attached where there is one; candidates named as
+// unconfirmed, never recited as his work.
+//
+// One renderer for the voice reply and the web page, for the same reason
+// DayPlan.Spoken is built core-side: two formatters drift, and then she says
+// one order and shows another.
+func FormatRU(ranked []Ranked) string {
+	var open, cands []Ranked
+	for _, r := range ranked {
+		if r.Status == StatusCandidate {
+			cands = append(cands, r)
+		} else {
+			open = append(open, r)
+		}
+	}
+	if len(open) == 0 && len(cands) == 0 {
+		return "задач нет."
+	}
+
+	var b strings.Builder
+	if len(open) > 0 {
+		b.WriteString("сначала: ")
+		b.WriteString(joinRU(open, SpokenLimit, true))
+		b.WriteString(".")
+	}
+	if len(cands) > 0 {
+		if b.Len() > 0 {
+			b.WriteString(" ")
+		}
+		b.WriteString("ещё я нашла, но ты не подтвердил: ")
+		b.WriteString(joinRU(cands, SpokenLimit, false))
+		b.WriteString(".")
+	}
+	return b.String()
+}
+
+// joinRU lists up to limit tasks, then says how many are left. withReasons
+// attaches the parenthesised reason — candidates are listed bare, since their
+// due dates are Maven's reading of a mail and not something he stated.
+func joinRU(rs []Ranked, limit int, withReasons bool) string {
+	shown := rs
+	rest := 0
+	if len(rs) > limit {
+		shown, rest = rs[:limit], len(rs)-limit
+	}
+	parts := make([]string, 0, len(shown))
+	for _, r := range shown {
+		if withReasons && r.Reason != "" {
+			parts = append(parts, r.Text+" ("+r.Reason+")")
+		} else {
+			parts = append(parts, r.Text)
+		}
+	}
+	s := strings.Join(parts, "; ")
+	if rest > 0 {
+		s += fmt.Sprintf("; и ещё %d", rest)
+	}
+	return s
+}

internal/tasks/rank_test.go

@@ -0,0 +1,177 @@
+package tasks
+
+import (
+	"strings"
+	"testing"
+	"time"
+)
+
+func at(y int, m time.Month, d int) *time.Time {
+	t := time.Date(y, m, d, 0, 0, 0, 0, time.UTC)
+	return &t
+}
+
+func now() time.Time { return time.Date(2026, 8, 1, 14, 0, 0, 0, time.UTC) }
+
+func texts(rs []Ranked) []string {
+	out := make([]string, len(rs))
+	for i, r := range rs {
+		out[i] = r.Text
+	}
+	return out
+}
+
+func TestRankOrdersByDeadline(t *testing.T) {
+	items := []Item{
+		{ID: 1, Text: "через неделю", Status: StatusOpen, Due: at(2026, 8, 7), Created: now()},
+		{ID: 2, Text: "просрочено", Status: StatusOpen, Due: at(2026, 7, 28), Created: now()},
+		{ID: 3, Text: "без срока", Status: StatusOpen, Created: now()},
+		{ID: 4, Text: "сегодня", Status: StatusOpen, Due: at(2026, 8, 1), Created: now()},
+		{ID: 5, Text: "завтра", Status: StatusOpen, Due: at(2026, 8, 2), Created: now()},
+	}
+	got := texts(Rank(items, now()))
+	want := []string{"просрочено", "сегодня", "завтра", "через неделю", "без срока"}
+	for i := range want {
+		if got[i] != want[i] {
+			t.Fatalf("order = %v, want %v", got, want)
+		}
+	}
+}
+
+func TestRankCandidatesNeverOutrankOpenWork(t *testing.T) {
+	items := []Item{
+		{ID: 1, Text: "его задача", Status: StatusOpen, Created: now()},
+		// Everything about this one screams urgent — and it is still a guess.
+		{ID: 2, Text: "из письма", Status: StatusCandidate, Due: at(2026, 7, 1), Weight: 3, Created: now()},
+	}
+	got := Rank(items, now())
+	if got[0].Text != "его задача" {
+		t.Errorf("order = %v, want his own work first", texts(got))
+	}
+}
+
+func TestRankWeightLiftsUndatedWork(t *testing.T) {
+	items := []Item{
+		{ID: 1, Text: "обычная", Status: StatusOpen, Created: now()},
+		{ID: 2, Text: "важная", Status: StatusOpen, Weight: 2, Created: now()},
+	}
+	got := Rank(items, now())
+	if got[0].Text != "важная" {
+		t.Errorf("order = %v, want the weighted task first", texts(got))
+	}
+	if got[0].Reason != "важно" {
+		t.Errorf("reason = %q, want важно", got[0].Reason)
+	}
+	// A deadline still beats a weight: a date is a fact, a weight is a feeling.
+	items = append(items, Item{ID: 3, Text: "сегодня", Status: StatusOpen, Due: at(2026, 8, 1), Created: now()})
+	got = Rank(items, now())
+	if got[0].Text != "сегодня" {
+		t.Errorf("order = %v, want the dated task first", texts(got))
+	}
+}
+
+func TestRankOldestFirstOnATie(t *testing.T) {
+	old := now().AddDate(0, 0, -3)
+	items := []Item{
+		{ID: 1, Text: "новая", Status: StatusOpen, Created: now()},
+		{ID: 2, Text: "старая", Status: StatusOpen, Created: old},
+	}
+	got := Rank(items, now())
+	if got[0].Text != "старая" {
+		t.Errorf("order = %v, want FIFO on equal urgency", texts(got))
+	}
+}
+
+func TestRankNoInventedReason(t *testing.T) {
+	got := Rank([]Item{{ID: 1, Text: "что-то", Status: StatusOpen, Created: now()}}, now())
+	if got[0].Reason != "" {
+		t.Errorf("reason = %q — nothing distinguished this task, so there is nothing to say", got[0].Reason)
+	}
+	if got[0].Score != 0 {
+		t.Errorf("score = %v, want 0", got[0].Score)
+	}
+}
+
+func TestRankAgeIsCappedAndNamed(t *testing.T) {
+	items := []Item{
+		{ID: 1, Text: "прошлогодняя", Status: StatusOpen, Created: now().AddDate(-1, 0, 0)},
+		{ID: 2, Text: "трёхнедельная", Status: StatusOpen, Created: now().AddDate(0, 0, -21)},
+	}
+	got := Rank(items, now())
+	if got[0].Score != scoreAgeCap {
+		t.Errorf("oldest score = %v, want the cap %v", got[0].Score, float64(scoreAgeCap))
+	}
+	if got[0].Reason != "давно в списке" {
+		t.Errorf("reason = %q", got[0].Reason)
+	}
+}
+
+// A task due at 23:00 today is due today, not overdue since this morning.
+func TestRankDueTodayIsNotOverdue(t *testing.T) {
+	due := time.Date(2026, 8, 1, 23, 0, 0, 0, time.UTC)
+	got := Rank([]Item{{ID: 1, Text: "вечером", Status: StatusOpen, Due: &due, Created: now()}}, now())
+	if got[0].Reason != "сегодня" {
+		t.Errorf("reason = %q, want сегодня", got[0].Reason)
+	}
+}
+
+func TestRankOverdueDaysAreCounted(t *testing.T) {
+	got := Rank([]Item{
+		{ID: 1, Text: "вчера", Status: StatusOpen, Due: at(2026, 7, 31), Created: now()},
+		{ID: 2, Text: "давно", Status: StatusOpen, Due: at(2026, 7, 20), Created: now()},
+	}, now())
+	if got[0].Text != "давно" {
+		t.Errorf("order = %v, want the later-overdue task first", texts(got))
+	}
+	if got[0].Reason != "просрочено на 12 дн." {
+		t.Errorf("reason = %q", got[0].Reason)
+	}
+	if got[1].Reason != "просрочено на день" {
+		t.Errorf("reason = %q", got[1].Reason)
+	}
+}
+
+func TestFormatRUNamesReasonsAndSeparatesCandidates(t *testing.T) {
+	ranked := Rank([]Item{
+		{ID: 1, Text: "оплатить интернет", Status: StatusOpen, Due: at(2026, 8, 1), Created: now()},
+		{ID: 2, Text: "купить молоко", Status: StatusOpen, Created: now()},
+		{ID: 3, Text: "продлить страховку", Status: StatusCandidate, Due: at(2026, 7, 1), Created: now()},
+	}, now())
+	got := FormatRU(ranked)
+	if !strings.HasPrefix(got, "сначала: оплатить интернет (сегодня)") {
+		t.Errorf("reply = %q", got)
+	}
+	if !strings.Contains(got, "не подтвердил: продлить страховку") {
+		t.Errorf("candidate not named as unconfirmed: %q", got)
+	}
+	// A candidate's due date is Maven's reading of a mail, not his statement.
+	if strings.Contains(got, "продлить страховку (") {
+		t.Errorf("a candidate must be listed without a reason: %q", got)
+	}
+	// Persona: nothing masculine, no pet names, informal address only.
+	for _, bad := range []string{"рад ", "понял ", "милый", "дорогой", "вам", "ваши"} {
+		if strings.Contains(got, bad) {
+			t.Errorf("reply %q contains %q", got, bad)
+		}
+	}
+}
+
+func TestFormatRUCapsTheSpokenList(t *testing.T) {
+	var items []Item
+	for i := 0; i < SpokenLimit+3; i++ {
+		items = append(items, Item{ID: int64(i), Text: "задача", Status: StatusOpen, Created: now()})
+	}
+	got := FormatRU(Rank(items, now()))
+	if !strings.Contains(got, "и ещё 3") {
+		t.Errorf("reply = %q, want the tail summarised", got)
+	}
+	if strings.Count(got, "задача") != SpokenLimit {
+		t.Errorf("reply = %q, want exactly %d named", got, SpokenLimit)
+	}
+}
+
+func TestFormatRUEmpty(t *testing.T) {
+	if got := FormatRU(nil); got != "задач нет." {
+		t.Errorf("reply = %q", got)
+	}
+}

internal/webfetch/webfetch.go

@@ -0,0 +1,301 @@
+// Package webfetch is the one door Maven uses to read something off the
+// network, and it is a narrow one.
+//
+// "Never phones home" stopped being a hard constraint on 2026-07-31, but what
+// replaced it is not "she may fetch anything": local sources come first (Kiwix
+// on the box), external fetching is off unless configured, and only the
+// utterance ever leaves — never his notes, facts or history. That policy is
+// enforced by the callers. What THIS package enforces is the part that must be
+// code rather than a paragraph in a plan, because it protects the homelab from
+// its own assistant:
+//
+//   - http/https only — no file://, no ftp://, no gopher;
+//   - no private address, ever: loopback, RFC1918 (which is what makes the
+//     10.42.0.0/24 wireguard tunnel and the 192.168.1.0/24 LAN unreachable),
+//     link-local incl. the 169.254.169.254 cloud metadata address, CGNAT,
+//     unique-local v6. Checked in the dialer's Control hook, so it holds for
+//     every address the resolver returns AND for every hop of a redirect
+//     chain — a DNS name that resolves to 127.0.0.1 is refused at connect
+//     time, which a pre-flight lookup could not promise (rebinding);
+//   - an allowlist, when one is configured, and a denylist that always wins;
+//   - a response size cap, a total timeout, a redirect cap;
+//   - one request per host per interval, so a poll loop with a bug is slow
+//     rather than an outbound flood.
+//
+// Everything above is on by default with sane numbers: a zero Config is a
+// usable, conservative fetcher. There is no cache and no retry — a feed poll
+// or a page read that fails is simply not answered this round.
+package webfetch
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"net"
+	"net/http"
+	"net/url"
+	"strings"
+	"sync"
+	"syscall"
+	"time"
+)
+
+// Defaults. Small on purpose: this reads feeds and article pages, not ISOs.
+const (
+	DefaultTimeout      = 20 * time.Second
+	DefaultMaxBytes     = 2 << 20 // 2 MiB
+	DefaultMaxRedirects = 3
+	DefaultHostInterval = time.Second
+	DefaultUserAgent    = "Maven/1.0 (self-hosted personal assistant)"
+)
+
+// Errors callers distinguish. Everything else is wrapped transport error.
+var (
+	ErrScheme    = errors.New("webfetch: only http and https are allowed")
+	ErrBlocked   = errors.New("webfetch: host is not allowed")
+	ErrPrivate   = errors.New("webfetch: refusing to connect to a private address")
+	ErrTooLarge  = errors.New("webfetch: response exceeds the size cap")
+	ErrRedirects = errors.New("webfetch: too many redirects")
+	ErrStatus    = errors.New("webfetch: non-2xx status")
+)
+
+// Config are the limits. Every zero value means "the default above", so
+// Config{} is safe; the only field that changes behaviour by being empty is
+// AllowHosts (empty ⇒ any public host that is not denied).
+type Config struct {
+	// AllowHosts — when non-empty, the ONLY hosts that may be fetched. An
+	// entry matches the host itself and its subdomains ("example.com" allows
+	// "news.example.com"). This is the knob to reach for when a capability
+	// should read two feeds and nothing else.
+	AllowHosts []string
+	// DenyHosts — same matching, checked first and always winning.
+	DenyHosts []string
+
+	Timeout      time.Duration // whole request, including redirects and body read
+	MaxBytes     int64         // response body cap
+	MaxRedirects int           // 0 ⇒ default; negative ⇒ no redirects followed
+	HostInterval time.Duration // minimum spacing between requests to one host
+	UserAgent    string
+
+	// AllowPrivate disables the private-address guard. It exists for tests
+	// (httptest listens on 127.0.0.1) and for an explicitly configured
+	// on-box mirror. Nothing in deploy/mavend.json sets it, and it should
+	// stay that way: with it on, any URL Maven is handed becomes an SSRF
+	// probe of the LAN and the wireguard range.
+	AllowPrivate bool
+}
+
+// Response is a fetched body, already bounded by MaxBytes.
+type Response struct {
+	URL         string // final URL after redirects
+	Status      int
+	ContentType string
+	Body        []byte
+}
+
+// Fetcher performs guarded GETs. Safe for concurrent use; the per-host rate
+// limiter is shared, which is the point of sharing one Fetcher.
+type Fetcher struct {
+	cfg  Config
+	http *http.Client
+
+	mu   sync.Mutex
+	last map[string]time.Time // host → when we last dialed it
+}
+
+// New builds a fetcher from cfg, filling in defaults.
+func New(cfg Config) *Fetcher {
+	if cfg.Timeout <= 0 {
+		cfg.Timeout = DefaultTimeout
+	}
+	if cfg.MaxBytes <= 0 {
+		cfg.MaxBytes = DefaultMaxBytes
+	}
+	if cfg.MaxRedirects == 0 {
+		cfg.MaxRedirects = DefaultMaxRedirects
+	}
+	if cfg.HostInterval <= 0 {
+		cfg.HostInterval = DefaultHostInterval
+	}
+	if cfg.UserAgent == "" {
+		cfg.UserAgent = DefaultUserAgent
+	}
+	f := &Fetcher{cfg: cfg, last: map[string]time.Time{}}
+
+	dialer := &net.Dialer{Timeout: 10 * time.Second}
+	if !cfg.AllowPrivate {
+		// The guard lives here rather than in a pre-flight net.LookupHost so
+		// that it sees the address actually being connected to: every A/AAAA
+		// the resolver handed back, on every redirect hop, with no window in
+		// which the name could be re-pointed at the LAN.
+		dialer.Control = func(_, address string, _ syscall.RawConn) error {
+			host, _, err := net.SplitHostPort(address)
+			if err != nil {
+				return err
+			}
+			ip := net.ParseIP(host)
+			if ip == nil || IsPrivateIP(ip) {
+				return fmt.Errorf("%w: %s", ErrPrivate, host)
+			}
+			return nil
+		}
+	}
+	f.http = &http.Client{
+		Timeout:   cfg.Timeout,
+		Transport: &http.Transport{DialContext: dialer.DialContext},
+		CheckRedirect: func(req *http.Request, via []*http.Request) error {
+			if len(via) > f.cfg.MaxRedirects {
+				return ErrRedirects
+			}
+			// A redirect is a fresh URL and gets the full check: an allowed
+			// host must not be able to bounce us onto a denied one.
+			return f.checkURL(req.URL)
+		},
+	}
+	return f
+}
+
+// Get fetches rawURL. The body is capped: a larger response is an error, not a
+// truncation, because half an XML document is worse than none.
+func (f *Fetcher) Get(ctx context.Context, rawURL string) (*Response, error) {
+	u, err := url.Parse(strings.TrimSpace(rawURL))
+	if err != nil {
+		return nil, fmt.Errorf("webfetch: bad url %q: %w", rawURL, err)
+	}
+	if err := f.checkURL(u); err != nil {
+		return nil, err
+	}
+	if err := f.waitTurn(ctx, u.Hostname()); err != nil {
+		return nil, err
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, u.String(), nil)
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("User-Agent", f.cfg.UserAgent)
+	req.Header.Set("Accept-Encoding", "identity")
+
+	resp, err := f.http.Do(req)
+	if err != nil {
+		// http.Client wraps our sentinels in *url.Error; unwrap so callers can
+		// still tell "blocked" from "the network is down".
+		for _, sentinel := range []error{ErrPrivate, ErrBlocked, ErrRedirects, ErrScheme} {
+			if errors.Is(err, sentinel) {
+				return nil, err
+			}
+		}
+		return nil, err
+	}
+	defer resp.Body.Close()
+
+	body, err := io.ReadAll(io.LimitReader(resp.Body, f.cfg.MaxBytes+1))
+	if err != nil {
+		return nil, err
+	}
+	if int64(len(body)) > f.cfg.MaxBytes {
+		return nil, fmt.Errorf("%w (%d bytes)", ErrTooLarge, f.cfg.MaxBytes)
+	}
+	if resp.StatusCode < 200 || resp.StatusCode > 299 {
+		return nil, fmt.Errorf("%w: %d", ErrStatus, resp.StatusCode)
+	}
+	return &Response{
+		URL:         resp.Request.URL.String(),
+		Status:      resp.StatusCode,
+		ContentType: resp.Header.Get("Content-Type"),
+		Body:        body,
+	}, nil
+}
+
+// checkURL applies the scheme rule and the host lists. The address rule is the
+// dialer's job (see New).
+func (f *Fetcher) checkURL(u *url.URL) error {
+	switch u.Scheme {
+	case "http", "https":
+	default:
+		return fmt.Errorf("%w: %q", ErrScheme, u.Scheme)
+	}
+	host := strings.ToLower(u.Hostname())
+	if host == "" {
+		return fmt.Errorf("%w: no host", ErrBlocked)
+	}
+	if HostMatches(host, f.cfg.DenyHosts) {
+		return fmt.Errorf("%w: %s is denied", ErrBlocked, host)
+	}
+	if len(f.cfg.AllowHosts) > 0 && !HostMatches(host, f.cfg.AllowHosts) {
+		return fmt.Errorf("%w: %s is not on the allowlist", ErrBlocked, host)
+	}
+	// A literal private address is refused here as well as in the dialer, so
+	// the error is the specific one even when no connection is attempted.
+	if !f.cfg.AllowPrivate {
+		if ip := net.ParseIP(host); ip != nil && IsPrivateIP(ip) {
+			return fmt.Errorf("%w: %s", ErrPrivate, host)
+		}
+	}
+	return nil
+}
+
+// waitTurn blocks until this host's rate-limit interval has elapsed. It holds
+// no lock while sleeping, so two hosts never wait on each other.
+func (f *Fetcher) waitTurn(ctx context.Context, host string) error {
+	for {
+		f.mu.Lock()
+		now := time.Now()
+		earliest := f.last[host].Add(f.cfg.HostInterval)
+		if !now.Before(earliest) {
+			f.last[host] = now
+			f.mu.Unlock()
+			return nil
+		}
+		f.mu.Unlock()
+		wait := time.NewTimer(earliest.Sub(now))
+		select {
+		case <-ctx.Done():
+			wait.Stop()
+			return ctx.Err()
+		case <-wait.C:
+		}
+	}
+}
+
+// HostMatches reports whether host equals one of pats or is a subdomain of one.
+// Exported because the crawler applies the same rule to links it decides not to
+// follow, before it ever builds a request.
+func HostMatches(host string, pats []string) bool {
+	host = strings.ToLower(strings.TrimSuffix(host, "."))
+	for _, p := range pats {
+		p = strings.ToLower(strings.TrimSpace(strings.TrimPrefix(p, "*.")))
+		if p == "" {
+			continue
+		}
+		if host == p || strings.HasSuffix(host, "."+p) {
+			return true
+		}
+	}
+	return false
+}
+
+// cgnat is 100.64.0.0/10 — carrier NAT, not covered by net.IP's helpers and not
+// somewhere a personal assistant has business connecting.
+var cgnat = &net.IPNet{IP: net.IPv4(100, 64, 0, 0).To4(), Mask: net.CIDRMask(10, 32)}
+
+// IsPrivateIP reports whether ip is somewhere Maven must never reach out to:
+// the box itself, the LAN, the wireguard range (10.42.0.0/24 ⊂ 10/8), the cloud
+// metadata address (169.254.169.254 ⊂ link-local), or anything unroutable.
+func IsPrivateIP(ip net.IP) bool {
+	if ip.IsLoopback() || ip.IsPrivate() || ip.IsUnspecified() ||
+		ip.IsLinkLocalUnicast() || ip.IsLinkLocalMulticast() ||
+		ip.IsInterfaceLocalMulticast() || ip.IsMulticast() {
+		return true
+	}
+	if v4 := ip.To4(); v4 != nil && cgnat.Contains(v4) {
+		return true
+	}
+	// IPv4-mapped/compatible forms of the above are handled by To4() inside the
+	// stdlib helpers; what is left is v6 unique-local (fc00::/7).
+	if len(ip) == net.IPv6len && ip.To4() == nil && ip[0]&0xfe == 0xfc {
+		return true
+	}
+	return false
+}

internal/webfetch/webfetch_test.go

@@ -0,0 +1,227 @@
+package webfetch
+
+import (
+	"context"
+	"errors"
+	"net"
+	"net/http"
+	"net/http/httptest"
+	"strings"
+	"testing"
+	"time"
+)
+
+// The limits in this package are the reason a crawler is allowed to exist on
+// this box at all, so each one has a test that fails loudly if it is removed.
+
+func TestPrivateAddressesAreRefused(t *testing.T) {
+	// The wireguard range (10.42.0.0/24), the LAN (192.168.1.0/24) and the
+	// cloud metadata address are the three that matter here; the rest come
+	// along for free.
+	for _, s := range []string{
+		"127.0.0.1", "127.1.2.3", "10.42.0.7", "10.0.0.5", "192.168.1.104",
+		"172.16.4.4", "169.254.169.254", "100.64.1.1", "0.0.0.0",
+		"::1", "fc00::1", "fd12:3456::1", "fe80::1",
+	} {
+		if !IsPrivateIP(net.ParseIP(s)) {
+			t.Errorf("IsPrivateIP(%s) = false, want true", s)
+		}
+	}
+	for _, s := range []string{"8.8.8.8", "1.1.1.1", "93.184.216.34", "2606:2800:220:1::1"} {
+		if IsPrivateIP(net.ParseIP(s)) {
+			t.Errorf("IsPrivateIP(%s) = true, want false", s)
+		}
+	}
+}
+
+func TestGetRefusesPrivateLiteral(t *testing.T) {
+	f := New(Config{})
+	for _, u := range []string{
+		"http://127.0.0.1:8034/search",
+		"http://10.42.0.1/",
+		"http://192.168.1.104/dash",
+		"http://[::1]:9100/mcp",
+	} {
+		if _, err := f.Get(context.Background(), u); !errors.Is(err, ErrPrivate) {
+			t.Errorf("Get(%s) error = %v, want ErrPrivate", u, err)
+		}
+	}
+}
+
+// A hostname that resolves into private space must fail too — that is the
+// rebinding case, and it is why the check lives in the dialer.
+func TestGetRefusesPrivateResolution(t *testing.T) {
+	f := New(Config{})
+	if _, err := f.Get(context.Background(), "http://localhost:8034/"); !errors.Is(err, ErrPrivate) {
+		t.Fatalf("Get(localhost) error = %v, want ErrPrivate", err)
+	}
+}
+
+func TestGetRefusesNonHTTPSchemes(t *testing.T) {
+	f := New(Config{})
+	for _, u := range []string{"file:///etc/passwd", "ftp://example.com/x", "gopher://example.com"} {
+		if _, err := f.Get(context.Background(), u); !errors.Is(err, ErrScheme) {
+			t.Errorf("Get(%s) error = %v, want ErrScheme", u, err)
+		}
+	}
+}
+
+// testFetcher — a fetcher pointed at an httptest server, which necessarily
+// listens on loopback. AllowPrivate is the test-only escape hatch.
+func testFetcher(t *testing.T, cfg Config) *Fetcher {
+	t.Helper()
+	cfg.AllowPrivate = true
+	if cfg.HostInterval == 0 {
+		cfg.HostInterval = time.Nanosecond
+	}
+	return New(cfg)
+}
+
+func TestAllowAndDenyLists(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Write([]byte("ok"))
+	}))
+	defer srv.Close()
+
+	f := testFetcher(t, Config{AllowHosts: []string{"example.com"}})
+	if _, err := f.Get(context.Background(), srv.URL); !errors.Is(err, ErrBlocked) {
+		t.Fatalf("off-allowlist host: error = %v, want ErrBlocked", err)
+	}
+	f = testFetcher(t, Config{DenyHosts: []string{"127.0.0.1"}})
+	if _, err := f.Get(context.Background(), srv.URL); !errors.Is(err, ErrBlocked) {
+		t.Fatalf("denied host: error = %v, want ErrBlocked", err)
+	}
+	f = testFetcher(t, Config{AllowHosts: []string{"127.0.0.1"}})
+	if _, err := f.Get(context.Background(), srv.URL); err != nil {
+		t.Fatalf("allowlisted host: %v", err)
+	}
+}
+
+func TestHostMatchesSubdomains(t *testing.T) {
+	pats := []string{"example.com", "*.news.org"}
+	for _, h := range []string{"example.com", "news.example.com", "a.b.example.com", "news.org", "feeds.news.org"} {
+		if !HostMatches(h, pats) {
+			t.Errorf("HostMatches(%q) = false, want true", h)
+		}
+	}
+	for _, h := range []string{"notexample.com", "example.com.evil.net", "org"} {
+		if HostMatches(h, pats) {
+			t.Errorf("HostMatches(%q) = true, want false", h)
+		}
+	}
+}
+
+func TestSizeCap(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Write([]byte(strings.Repeat("x", 5000)))
+	}))
+	defer srv.Close()
+
+	f := testFetcher(t, Config{MaxBytes: 100})
+	if _, err := f.Get(context.Background(), srv.URL); !errors.Is(err, ErrTooLarge) {
+		t.Fatalf("error = %v, want ErrTooLarge", err)
+	}
+	f = testFetcher(t, Config{MaxBytes: 6000})
+	resp, err := f.Get(context.Background(), srv.URL)
+	if err != nil {
+		t.Fatalf("under the cap: %v", err)
+	}
+	if len(resp.Body) != 5000 {
+		t.Fatalf("body = %d bytes, want 5000", len(resp.Body))
+	}
+}
+
+func TestRedirectCap(t *testing.T) {
+	var srv *httptest.Server
+	srv = httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		http.Redirect(w, r, srv.URL+"/again", http.StatusFound)
+	}))
+	defer srv.Close()
+
+	f := testFetcher(t, Config{MaxRedirects: 2})
+	if _, err := f.Get(context.Background(), srv.URL); !errors.Is(err, ErrRedirects) {
+		t.Fatalf("error = %v, want ErrRedirects", err)
+	}
+}
+
+// A redirect off the allowlist is the interesting redirect: the first hop is
+// permitted, the second must not be.
+func TestRedirectRecheckedAgainstDenylist(t *testing.T) {
+	target := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Write([]byte("secret"))
+	}))
+	defer target.Close()
+	hop := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		http.Redirect(w, r, target.URL, http.StatusFound)
+	}))
+	defer hop.Close()
+
+	// Reach the hop under the name "localhost" and allow only that name; the
+	// redirect lands on the same box under its literal address, which the
+	// allowlist does not cover. Without the CheckRedirect hook this fetch
+	// succeeds and returns "secret".
+	f := testFetcher(t, Config{AllowHosts: []string{"localhost"}})
+	viaName := strings.Replace(hop.URL, "127.0.0.1", "localhost", 1)
+	if _, err := f.Get(context.Background(), viaName); !errors.Is(err, ErrBlocked) {
+		t.Fatalf("error = %v, want ErrBlocked", err)
+	}
+}
+
+func TestPerHostRateLimit(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Write([]byte("ok"))
+	}))
+	defer srv.Close()
+
+	f := testFetcher(t, Config{HostInterval: 60 * time.Millisecond})
+	start := time.Now()
+	for i := 0; i < 3; i++ {
+		if _, err := f.Get(context.Background(), srv.URL); err != nil {
+			t.Fatalf("request %d: %v", i, err)
+		}
+	}
+	if elapsed := time.Since(start); elapsed < 120*time.Millisecond {
+		t.Fatalf("three requests took %s, want at least 120ms of spacing", elapsed)
+	}
+}
+
+func TestRateLimitHonoursContext(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {}))
+	defer srv.Close()
+
+	f := testFetcher(t, Config{HostInterval: 10 * time.Second})
+	if _, err := f.Get(context.Background(), srv.URL); err != nil {
+		t.Fatalf("first request: %v", err)
+	}
+	ctx, cancel := context.WithTimeout(context.Background(), 20*time.Millisecond)
+	defer cancel()
+	if _, err := f.Get(ctx, srv.URL); !errors.Is(err, context.DeadlineExceeded) {
+		t.Fatalf("error = %v, want DeadlineExceeded", err)
+	}
+}
+
+func TestNon2xxIsAnError(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		http.Error(w, "nope", http.StatusInternalServerError)
+	}))
+	defer srv.Close()
+	f := testFetcher(t, Config{})
+	if _, err := f.Get(context.Background(), srv.URL); !errors.Is(err, ErrStatus) {
+		t.Fatalf("error = %v, want ErrStatus", err)
+	}
+}
+
+func TestUserAgentIsSent(t *testing.T) {
+	got := make(chan string, 1)
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		got <- r.Header.Get("User-Agent")
+	}))
+	defer srv.Close()
+	f := testFetcher(t, Config{UserAgent: "Maven/test"})
+	if _, err := f.Get(context.Background(), srv.URL); err != nil {
+		t.Fatal(err)
+	}
+	if ua := <-got; ua != "Maven/test" {
+		t.Fatalf("user-agent = %q", ua)
+	}
+}

internal/zenmoney/client.go

@@ -0,0 +1,267 @@
+// Package zenmoney reads spending and income from ZenMoney's /v8/diff/ API
+// (Vikunja #125).
+//
+// Trust boundary: ZenMoney, not Maven. They already hold his bank sessions —
+// this package only reads back what they have, over a token that lives in the
+// poller module and is never handed to core. Nothing here writes to ZenMoney;
+// diff is called read-only (an empty change set in, a change set out).
+//
+// Two rules the code exists to enforce:
+//
+//   - NEVER invent a number. Every figure in a Summary is a sum of amounts the
+//     API returned. A request that fails, or returns nothing, produces no
+//     summary and therefore no fact — silence, not a zero. A confidently wrong
+//     "ты потратил 0" is worse than no answer.
+//   - His money is never search input. This package holds no notes, no
+//     utterances and no persona text, and it has no path to the external search
+//     capability. The only thing that leaves the box here is the diff request
+//     itself, to the service that already has the data.
+package zenmoney
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"fmt"
+	"io"
+	"net/http"
+	"sort"
+	"strings"
+	"time"
+)
+
+// DefaultBaseURL — ZenMoney's API root. Overridable so the tests can point at
+// an httptest server replaying a recorded response.
+const DefaultBaseURL = "https://api.zenmoney.ru"
+
+// Client is a ZenMoney diff reader. The token is held here, in the poller's
+// address space; core never receives it and never learns it exists.
+type Client struct {
+	BaseURL string
+	Token   string
+	HTTP    *http.Client
+}
+
+// New returns a client with a bounded HTTP timeout. An empty token is a
+// programming error the caller must catch — the capability is off unless
+// configured, so a client is only ever built when a token was supplied.
+func New(token, baseURL string, timeout time.Duration) (*Client, error) {
+	if strings.TrimSpace(token) == "" {
+		return nil, fmt.Errorf("zenmoney: empty token")
+	}
+	if baseURL == "" {
+		baseURL = DefaultBaseURL
+	}
+	if timeout <= 0 {
+		timeout = 20 * time.Second
+	}
+	return &Client{
+		BaseURL: strings.TrimRight(baseURL, "/"),
+		Token:   token,
+		HTTP:    &http.Client{Timeout: timeout},
+	}, nil
+}
+
+// diffRequest — the smallest body /v8/diff/ accepts. serverTimestamp is the
+// incremental cursor: the server returns objects changed at or after it.
+type diffRequest struct {
+	CurrentClientTimestamp int64 `json:"currentClientTimestamp"`
+	ServerTimestamp        int64 `json:"serverTimestamp"`
+}
+
+// diffResponse — only the fields spending needs. ZenMoney returns a dozen more
+// object types (tags, merchants, budgets, reminders); decoding them would mean
+// holding more of his financial life in memory than the question needs.
+type diffResponse struct {
+	ServerTimestamp int64         `json:"serverTimestamp"`
+	Instrument      []instrument  `json:"instrument"`
+	Transaction     []transaction `json:"transaction"`
+}
+
+type instrument struct {
+	ID         int64  `json:"id"`
+	ShortTitle string `json:"shortTitle"`
+}
+
+type transaction struct {
+	ID               string  `json:"id"`
+	Date             string  `json:"date"` // "2026-07-15"
+	Deleted          bool    `json:"deleted"`
+	Income           float64 `json:"income"`
+	Outcome          float64 `json:"outcome"`
+	IncomeInstrument int64   `json:"incomeInstrument"`
+	OutcomeInstrmnt  int64   `json:"outcomeInstrument"`
+	IncomeAccount    string  `json:"incomeAccount"`
+	OutcomeAccount   string  `json:"outcomeAccount"`
+}
+
+// Money — an amount in one currency. Kept as the currency's own short title
+// ("RUB", "EUR") rather than converted: ZenMoney's rates are a snapshot, and
+// converting would turn a figure he can check against his bank into one he
+// cannot.
+type Money struct {
+	Currency string  `json:"currency"`
+	Amount   float64 `json:"amount"`
+}
+
+// Summary — what was spent and earned over a window, per currency, plus how
+// many transactions it was computed from. Count is the honesty check: a
+// summary built from zero transactions is not "you spent nothing", it is "there
+// was nothing to read", and callers treat it as no answer.
+type Summary struct {
+	From, To time.Time
+	Spent    []Money `json:"spent"`
+	Earned   []Money `json:"earned"`
+	Count    int     `json:"count"`
+	// ServerTimestamp — the cursor the API returned, for the caller to log or
+	// carry. Not used as an incremental cursor for summaries; see Since.
+	ServerTimestamp int64 `json:"-"`
+}
+
+// Since returns the summary of transactions dated in [from, to).
+//
+// The diff cursor is set to `from` so the server only sends objects changed
+// since then, which for a "this month" window is everything filed this month.
+// The caveat, deliberately accepted: a transaction he EDITED this month but
+// dated last month arrives too, and is then excluded by date — so editing old
+// records cannot inflate this month's total. The reverse case (a transaction
+// dated this month, filed and last changed before `from`) cannot exist.
+func (c *Client) Since(ctx context.Context, from, to time.Time) (Summary, error) {
+	resp, err := c.diff(ctx, from.Unix())
+	if err != nil {
+		return Summary{}, err
+	}
+	return summarize(resp, from, to), nil
+}
+
+func (c *Client) diff(ctx context.Context, serverTimestamp int64) (diffResponse, error) {
+	body, err := json.Marshal(diffRequest{
+		CurrentClientTimestamp: time.Now().Unix(),
+		ServerTimestamp:        serverTimestamp,
+	})
+	if err != nil {
+		return diffResponse{}, err
+	}
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, c.BaseURL+"/v8/diff/", bytes.NewReader(body))
+	if err != nil {
+		return diffResponse{}, err
+	}
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Authorization", "Bearer "+c.Token)
+	hc := c.HTTP
+	if hc == nil {
+		hc = &http.Client{Timeout: 20 * time.Second}
+	}
+	res, err := hc.Do(req)
+	if err != nil {
+		return diffResponse{}, err
+	}
+	defer res.Body.Close()
+	raw, err := io.ReadAll(io.LimitReader(res.Body, 32<<20))
+	if err != nil {
+		return diffResponse{}, err
+	}
+	if res.StatusCode != http.StatusOK {
+		// The status only. The body of a failed diff can echo account data, and
+		// this string reaches the log.
+		return diffResponse{}, fmt.Errorf("zenmoney diff: %s", res.Status)
+	}
+	var out diffResponse
+	if err := json.Unmarshal(raw, &out); err != nil {
+		return diffResponse{}, fmt.Errorf("zenmoney diff: decode: %w", err)
+	}
+	return out, nil
+}
+
+// summarize sums the transactions dated inside the window.
+//
+// Excluded, in order: deleted rows (ZenMoney tombstones rather than removes),
+// transfers and currency exchanges (income and outcome both non-zero — moving
+// his own money between his own accounts is not spending), and anything dated
+// outside the window.
+func summarize(resp diffResponse, from, to time.Time) Summary {
+	cur := map[int64]string{}
+	for _, in := range resp.Instrument {
+		cur[in.ID] = in.ShortTitle
+	}
+	spent := map[string]float64{}
+	earned := map[string]float64{}
+	count := 0
+	for _, t := range resp.Transaction {
+		if t.Deleted {
+			continue
+		}
+		d, err := time.ParseInLocation("2006-01-02", t.Date, from.Location())
+		if err != nil {
+			continue // an undated row is not a number we can place
+		}
+		if d.Before(from) || !d.Before(to) {
+			continue
+		}
+		if t.Income > 0 && t.Outcome > 0 {
+			continue // transfer / exchange
+		}
+		switch {
+		case t.Outcome > 0:
+			spent[currency(cur, t.OutcomeInstrmnt)] += t.Outcome
+			count++
+		case t.Income > 0:
+			earned[currency(cur, t.IncomeInstrument)] += t.Income
+			count++
+		}
+	}
+	return Summary{
+		From: from, To: to,
+		Spent: sortMoney(spent), Earned: sortMoney(earned),
+		Count: count, ServerTimestamp: resp.ServerTimestamp,
+	}
+}
+
+// currency names the instrument, or says it does not know. An unknown id keeps
+// the amount rather than dropping it: a sum without a currency label is still
+// his money, and silently discarding it would understate the total.
+func currency(names map[int64]string, id int64) string {
+	if s := names[id]; s != "" {
+		return s
+	}
+	return "?"
+}
+
+// sortMoney gives the amounts a stable order (largest first) so the rendered
+// string and the written fact do not churn between polls.
+func sortMoney(m map[string]float64) []Money {
+	out := make([]Money, 0, len(m))
+	for c, a := range m {
+		out = append(out, Money{Currency: c, Amount: a})
+	}
+	sort.Slice(out, func(i, j int) bool {
+		if out[i].Amount != out[j].Amount {
+			return out[i].Amount > out[j].Amount
+		}
+		return out[i].Currency < out[j].Currency
+	})
+	return out
+}
+
+// Empty reports whether the summary rests on no transactions at all. Callers
+// must treat an empty summary as "nothing to say", never as a zero: the
+// difference between "he spent nothing" and "the read returned nothing" is the
+// difference between an answer and an invented one.
+func (s Summary) Empty() bool { return s.Count == 0 }
+
+// MonthWindow — the first instant of now's month, and now's own day-end
+// exclusive bound, in now's location. The window a "сколько я потратил в этом
+// месяце?" question means.
+func MonthWindow(now time.Time) (from, to time.Time) {
+	loc := now.Location()
+	from = time.Date(now.Year(), now.Month(), 1, 0, 0, 0, 0, loc)
+	to = time.Date(now.Year(), now.Month(), now.Day(), 0, 0, 0, 0, loc).AddDate(0, 0, 1)
+	return from, to
+}
+
+// DayWindow — today, in now's location.
+func DayWindow(now time.Time) (from, to time.Time) {
+	loc := now.Location()
+	from = time.Date(now.Year(), now.Month(), now.Day(), 0, 0, 0, 0, loc)
+	return from, from.AddDate(0, 0, 1)
+}

internal/zenmoney/client_test.go

@@ -0,0 +1,178 @@
+package zenmoney
+
+import (
+	"context"
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"os"
+	"strings"
+	"testing"
+	"time"
+)
+
+// fixtureServer replays testdata/diff.json and records the request, so the
+// tests can assert the wire contract (Bearer token, POST, /v8/diff/) without a
+// ZenMoney account.
+func fixtureServer(t *testing.T, got *diffRequest, auth *string) *httptest.Server {
+	t.Helper()
+	body, err := os.ReadFile("testdata/diff.json")
+	if err != nil {
+		t.Fatal(err)
+	}
+	return httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		if r.Method != http.MethodPost {
+			t.Errorf("method = %s, want POST", r.Method)
+		}
+		if r.URL.Path != "/v8/diff/" {
+			t.Errorf("path = %s, want /v8/diff/", r.URL.Path)
+		}
+		if auth != nil {
+			*auth = r.Header.Get("Authorization")
+		}
+		if got != nil {
+			if err := json.NewDecoder(r.Body).Decode(got); err != nil {
+				t.Errorf("decode request: %v", err)
+			}
+		}
+		w.Header().Set("Content-Type", "application/json")
+		w.Write(body)
+	}))
+}
+
+func aug(day int) time.Time { return time.Date(2026, 8, day, 0, 0, 0, 0, time.UTC) }
+
+func TestSinceSumsSpendingPerCurrency(t *testing.T) {
+	var req diffRequest
+	var auth string
+	srv := fixtureServer(t, &req, &auth)
+	defer srv.Close()
+
+	c, err := New("tok", srv.URL, time.Second)
+	if err != nil {
+		t.Fatal(err)
+	}
+	s, err := c.Since(context.Background(), aug(1), aug(6))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if auth != "Bearer tok" {
+		t.Errorf("Authorization = %q", auth)
+	}
+	if req.ServerTimestamp != aug(1).Unix() {
+		t.Errorf("serverTimestamp = %d, want the window start", req.ServerTimestamp)
+	}
+	// 1500 + 249.5 RUB spent, 12 EUR spent, 3000 RUB in. The transfer (t4), the
+	// deleted row (t6) and July's salary (t3) are all excluded.
+	want := map[string]float64{"RUB": 1749.5, "EUR": 12}
+	if len(s.Spent) != 2 {
+		t.Fatalf("spent = %+v, want two currencies", s.Spent)
+	}
+	for _, m := range s.Spent {
+		if want[m.Currency] != m.Amount {
+			t.Errorf("spent %s = %v, want %v", m.Currency, m.Amount, want[m.Currency])
+		}
+	}
+	if len(s.Earned) != 1 || s.Earned[0].Amount != 3000 || s.Earned[0].Currency != "RUB" {
+		t.Errorf("earned = %+v, want 3000 RUB (July's salary is outside the window)", s.Earned)
+	}
+	if s.Count != 4 {
+		t.Errorf("count = %d, want 4 counted transactions", s.Count)
+	}
+	// Largest first, so the fact value does not churn between polls.
+	if s.Spent[0].Currency != "RUB" {
+		t.Errorf("spent order = %+v, want the largest amount first", s.Spent)
+	}
+}
+
+// A window with nothing in it is NOT a zero. No transactions means no answer,
+// and the caller must be able to tell the difference.
+func TestSinceEmptyWindowIsNotAZero(t *testing.T) {
+	srv := fixtureServer(t, nil, nil)
+	defer srv.Close()
+	c, _ := New("tok", srv.URL, time.Second)
+	s, err := c.Since(context.Background(), time.Date(2026, 9, 1, 0, 0, 0, 0, time.UTC), time.Date(2026, 9, 30, 0, 0, 0, 0, time.UTC))
+	if err != nil {
+		t.Fatal(err)
+	}
+	if !s.Empty() {
+		t.Fatalf("summary = %+v, want empty", s)
+	}
+	if _, ok := s.Value(); ok {
+		t.Error("an empty summary must not produce a fact value")
+	}
+}
+
+func TestSinceReportsHTTPFailureWithoutTheBody(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		http.Error(w, `{"account":"acc-card","secret":"leaky"}`, http.StatusUnauthorized)
+	}))
+	defer srv.Close()
+	c, _ := New("tok", srv.URL, time.Second)
+	_, err := c.Since(context.Background(), aug(1), aug(6))
+	if err == nil {
+		t.Fatal("want an error on 401")
+	}
+	if strings.Contains(err.Error(), "acc-card") || strings.Contains(err.Error(), "leaky") {
+		t.Errorf("error %q echoes the response body — it reaches the log", err)
+	}
+}
+
+func TestNewRequiresAToken(t *testing.T) {
+	if _, err := New("  ", "", 0); err == nil {
+		t.Error("want an error for an empty token — the capability is off unless configured")
+	}
+}
+
+func TestMonthAndDayWindows(t *testing.T) {
+	now := time.Date(2026, 8, 15, 21, 30, 0, 0, time.UTC)
+	from, to := MonthWindow(now)
+	if from != time.Date(2026, 8, 1, 0, 0, 0, 0, time.UTC) || to != time.Date(2026, 8, 16, 0, 0, 0, 0, time.UTC) {
+		t.Errorf("month window = %v..%v", from, to)
+	}
+	from, to = DayWindow(now)
+	if from != time.Date(2026, 8, 15, 0, 0, 0, 0, time.UTC) || to != time.Date(2026, 8, 16, 0, 0, 0, 0, time.UTC) {
+		t.Errorf("day window = %v..%v", from, to)
+	}
+}
+
+func TestFactValueRoundTripAndFormat(t *testing.T) {
+	s := Summary{Spent: []Money{{"RUB", 1749.5}}, Earned: []Money{{"RUB", 3000}}, Count: 3}
+	raw, ok := s.Value()
+	if !ok {
+		t.Fatal("want a fact value")
+	}
+	v, err := ParseFactValue(raw)
+	if err != nil {
+		t.Fatal(err)
+	}
+	got := v.FormatRU("в этом месяце")
+	if !strings.Contains(got, "1749.5 RUB") || !strings.Contains(got, "3000 RUB") {
+		t.Errorf("reply = %q, want the exact figures", got)
+	}
+	// Persona: informal, feminine, no commentary on his spending.
+	for _, bad := range []string{"вы", "ваш", "милый", "дорогой", "рад ", "слишком", "много"} {
+		if strings.Contains(got, bad) {
+			t.Errorf("reply %q contains %q", got, bad)
+		}
+	}
+	if strings.Contains(got, "он ") {
+		t.Errorf("reply %q talks about him in the third person", got)
+	}
+}
+
+// An empty fact value renders to nothing, so a caller cannot accidentally
+// speak a zero.
+func TestFormatRUEmptyRendersNothing(t *testing.T) {
+	if got := (FactValue{}).FormatRU("сегодня"); got != "" {
+		t.Errorf("reply = %q, want empty", got)
+	}
+}
+
+func TestFormatAmountKeepsTheTruth(t *testing.T) {
+	for in, want := range map[float64]string{1500: "1500", 249.5: "249.5", 0.99: "0.99", 1749.55: "1749.55"} {
+		if got := formatAmount(in); got != want {
+			t.Errorf("formatAmount(%v) = %q, want %q", in, got, want)
+		}
+	}
+}

internal/zenmoney/render.go

@@ -0,0 +1,99 @@
+package zenmoney
+
+import (
+	"encoding/json"
+	"fmt"
+	"strings"
+	"time"
+)
+
+// Fact keys the poller writes, all under source "poll:zenmoney". Two windows,
+// because they are the two questions he actually asks; a per-category
+// breakdown would mean storing what he bought, and the store is not a ledger.
+const (
+	KeySpentToday = "money_today"
+	KeySpentMonth = "money_month"
+)
+
+// Source — the provenance every money fact carries. The loop's rules trust
+// source, and nothing in Maven has a rule on these keys: they are read when he
+// asks, never a reason to speak. Maven is not a nag, least of all about money.
+const Source = "poll:zenmoney"
+
+// FactValue — the JSON stored in a money fact. A wire shape of its own rather
+// than the Summary struct so From/To (which carry a timezone and a clock) stay
+// out of the store; the key already says which window it is.
+type FactValue struct {
+	Spent  []Money `json:"spent"`
+	Earned []Money `json:"earned"`
+	Count  int     `json:"count"`
+}
+
+// Value encodes the summary for the facts table. Returns ok=false for an empty
+// summary: no transactions read means no fact written, so that a failed or
+// empty poll can never be recited back to him as a zero.
+func (s Summary) Value() (string, bool) {
+	if s.Empty() {
+		return "", false
+	}
+	b, err := json.Marshal(FactValue{Spent: s.Spent, Earned: s.Earned, Count: s.Count})
+	if err != nil {
+		return "", false
+	}
+	return string(b), true
+}
+
+// ParseFactValue decodes a stored money fact.
+func ParseFactValue(raw string) (FactValue, error) {
+	var v FactValue
+	if err := json.Unmarshal([]byte(raw), &v); err != nil {
+		return FactValue{}, err
+	}
+	return v, nil
+}
+
+// FormatRU renders a money fact the way Maven says it — feminine, informal,
+// and only about numbers that came from ZenMoney. window is the Russian phrase
+// for the period ("сегодня", "в этом месяце").
+//
+// No commentary. She reports the figure and stops: an opinion about his
+// spending is exactly the nagging Maven is not for.
+func (v FactValue) FormatRU(window string) string {
+	if v.Count == 0 {
+		return ""
+	}
+	var parts []string
+	if len(v.Spent) > 0 {
+		parts = append(parts, "потратил "+joinMoney(v.Spent))
+	}
+	if len(v.Earned) > 0 {
+		parts = append(parts, "получил "+joinMoney(v.Earned))
+	}
+	if len(parts) == 0 {
+		return ""
+	}
+	return window + " ты " + strings.Join(parts, ", ") + "."
+}
+
+func joinMoney(ms []Money) string {
+	parts := make([]string, 0, len(ms))
+	for _, m := range ms {
+		parts = append(parts, fmt.Sprintf("%s %s", formatAmount(m.Amount), m.Currency))
+	}
+	return strings.Join(parts, " и ")
+}
+
+// formatAmount — whole units when the amount is whole, two decimals otherwise.
+// Never rounded to something prettier than the truth.
+func formatAmount(a float64) string {
+	if a == float64(int64(a)) {
+		return fmt.Sprintf("%d", int64(a))
+	}
+	return strings.TrimRight(strings.TrimRight(fmt.Sprintf("%.2f", a), "0"), ".")
+}
+
+// StaleAfter — how old a money fact may be and still be worth reciting. The
+// poller is off unless configured and can be down; answering with last week's
+// total as if it were today's would be a lie by omission, so a stale fact is
+// reported as stale.
+const StaleAfter = 26 * time.Hour

internal/zenmoney/testdata/diff.json

@@ -0,0 +1,35 @@
+{
+  "serverTimestamp": 1785312000,
+  "instrument": [
+    {"id": 2, "title": "Российский рубль", "shortTitle": "RUB", "symbol": "₽", "rate": 1},
+    {"id": 3, "title": "Евро", "shortTitle": "EUR", "symbol": "€", "rate": 100}
+  ],
+  "account": [
+    {"id": "acc-card", "title": "карта", "instrument": 2},
+    {"id": "acc-cash", "title": "наличные", "instrument": 2},
+    {"id": "acc-eur", "title": "евро", "instrument": 3}
+  ],
+  "transaction": [
+    {"id": "t1", "date": "2026-08-01", "changed": 1785300000, "income": 0, "outcome": 1500,
+     "incomeInstrument": 2, "outcomeInstrument": 2, "incomeAccount": "acc-card", "outcomeAccount": "acc-card",
+     "payee": "пятёрочка", "deleted": false},
+    {"id": "t2", "date": "2026-08-01", "changed": 1785300001, "income": 0, "outcome": 249.5,
+     "incomeInstrument": 2, "outcomeInstrument": 2, "incomeAccount": "acc-card", "outcomeAccount": "acc-card",
+     "payee": "метро", "deleted": false},
+    {"id": "t3", "date": "2026-07-20", "changed": 1785300002, "income": 120000, "outcome": 0,
+     "incomeInstrument": 2, "outcomeInstrument": 2, "incomeAccount": "acc-card", "outcomeAccount": "acc-card",
+     "payee": "зарплата", "deleted": false},
+    {"id": "t4", "date": "2026-08-02", "changed": 1785300003, "income": 5000, "outcome": 5000,
+     "incomeInstrument": 2, "outcomeInstrument": 2, "incomeAccount": "acc-cash", "outcomeAccount": "acc-card",
+     "payee": "", "comment": "снял наличные", "deleted": false},
+    {"id": "t5", "date": "2026-08-03", "changed": 1785300004, "income": 0, "outcome": 12,
+     "incomeInstrument": 3, "outcomeInstrument": 3, "incomeAccount": "acc-eur", "outcomeAccount": "acc-eur",
+     "payee": "hosting", "deleted": false},
+    {"id": "t6", "date": "2026-08-04", "changed": 1785300005, "income": 0, "outcome": 999,
+     "incomeInstrument": 2, "outcomeInstrument": 2, "incomeAccount": "acc-card", "outcomeAccount": "acc-card",
+     "payee": "удалённая", "deleted": true},
+    {"id": "t7", "date": "2026-08-05", "changed": 1785300006, "income": 3000, "outcome": 0,
+     "incomeInstrument": 2, "outcomeInstrument": 2, "incomeAccount": "acc-card", "outcomeAccount": "acc-card",
+     "payee": "возврат", "deleted": false}
+  ]
+}