Switch menu source from Nutrislice to Claude-read PDF, weekly to monthly pull

The district discontinued Nutrislice in favor of a monthly PDF calendar
per grade band on kyrene.org. That PDF is a graphic calendar grid, not
reliably text-extractable, so scraper.py now sends it to Claude (vision +
structured JSON output) instead of hitting a JSON API. Fetches through a
stable Finalsite resource-manager permalink that always redirects to
whichever month is currently published, so no code change is needed each
month.

Also moves the pull schedule from weekly (day-of-week list) to monthly
(day-of-month), since the source itself is now monthly — defaults to the
25th, since the district tends to post next month's PDF a few days early.

Requires a new ANTHROPIC_API_KEY in the server's .env.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013RbBoYm23muVyMZ18Ldqa6

.env.example

@@ -6,12 +6,17 @@
 # e.g. http://192.168.86.X:8123/api/webhook/<random-token>
 HA_WEBHOOK_URL=
 
+# Anthropic API key. The district's lunch menu is now a graphic PDF
+# calendar (Nutrislice was discontinued), so scraper.py sends it to Claude
+# to read into structured data — see console.anthropic.com/settings/keys.
+ANTHROPIC_API_KEY=
+
 # --- Optional ---
-# Day(s) of week to pull the upcoming school week and push it to the
-# calendar (comma-separated: mon,tue,wed,thu,fri,sat,sun). Picking more
-# than one day creates duplicate events for days already pushed — most
-# people want just one.
-CALENDAR_RUN_DAYS=sun
+# Day of the month (1-28) to pull whichever month is currently published
+# and push it to the calendar. Re-pushing the same month creates duplicate
+# events, so pick one day near month-end (the district tends to post next
+# month's menu a few days early).
+CALENDAR_RUN_DAY_OF_MONTH=25
 
 # Time to run the pull (24-hour HH:MM, container timezone)
 CALENDAR_RUN_TIME=18:00

CLAUDE.md

@@ -1,9 +1,8 @@
 # Lunch Notifier
 
-Single-file Flask app that scrapes the weekly Mariposa (Kyrene school
-district) lunch menu and pushes each day to a Home Assistant calendar as an
-all-day event. Runs as one Docker container on the user's Unraid home
-server.
+Single-file Flask app that scrapes the Mariposa (Kyrene school district)
+lunch menu and pushes each day to a Home Assistant calendar as an all-day
+event. Runs as one Docker container on the user's Unraid home server.
 
 **2026-09-03: pivoted from SMS to calendar-only.** Originally texted phone
 numbers via email-to-SMS carrier gateways; that whole path (Gmail SMTP,
@@ -12,40 +11,70 @@ user's request in favor of Home Assistant calendar events exclusively. If
 you find references to texting/SMS/phone numbers anywhere (docs, old
 commits), they describe the pre-2026-09-03 app, not the current one.
 
+**2026-09-08: pivoted from Nutrislice to a Claude-read PDF, and from
+weekly to monthly pulls.** The district discontinued Nutrislice (the old
+scraper hit `kyrene.api.nutrislice.com`, a clean per-day JSON API) in
+favor of publishing one monthly menu PDF per grade band on
+`kyrene.org/family-student-resources/food-services/school-menus`. See
+"Menu source" and "Monthly pull scheduler" below for how that's handled
+now. Anything describing a weekly Nutrislice pull describes the
+pre-2026-09-08 app.
+
 ## Architecture
 
 - **`app.py`** — everything: Flask web UI/API, `schedule`-library job
   scheduling, and the Home Assistant webhook push. The HTML/CSS/JS frontend
   is embedded as a Python string constant (`HTML`) rather than a separate
   template file — no build step, no template engine.
-- **`scraper.py`** — fetches and parses the school's published lunch menu
-  from Nutrislice's API. `get_week_menu(date_str)` returns every day in the
-  API's week-granular response as `[{"date": ..., "items": [...]}, ...]`
-  (the API returns a full week regardless of which single date you ask
-  for). Days with no posted menu (weekends, breaks, holidays) come back
-  with an empty `items` list rather than being omitted.
+- **`scraper.py`** — fetches the currently-published K-5 lunch menu PDF and
+  asks Claude (Anthropic API, `claude-opus-5`, structured JSON output via
+  `output_config.format`) to read it into `[{"date": ..., "items": [...]}, ...]`
+  for every day of whatever month is published. Days with no posted menu
+  (weekends, breaks, holidays) come back with an empty `items` list rather
+  than being omitted — same contract the old Nutrislice-based scraper used,
+  so `app.py`'s push logic didn't need to change.
+- **Menu source (Nutrislice → PDF, 2026-09-08):** the district's page links
+  a *stable* Finalsite "resource manager" permalink
+  (`kyrene.org/fs/resource-manager/view/32bb3f8b-5a50-4b38-ad68-a5d3108de571`
+  for the K-5/"Standard" menu — there's a different UUID per grade band/
+  campus on that page) that 302-redirects to whichever monthly PDF is
+  currently live (`scraper.MENU_PAGE_URL`). **Don't hardcode the PDF URL
+  itself** — it's a Finalsite CDN path that changes every month (verified:
+  the September 2026 PDF and August's page both linked through the same
+  permalink); only the permalink UUID is stable, and only changes if the
+  district restructures that page entirely. The PDF itself is a single-page
+  graphic calendar grid (mascots, colored cells, day numbers laid out
+  visually) — not text-extractable in any reliable way (raw text extraction
+  jumbles day numbers away from their menu items), which is why this reads
+  it via Claude vision/document understanding rather than `pdfplumber`/
+  regex. Cost is negligible: one ~370KB single-page PDF per month, well
+  under a cent of Opus 5 tokens.
+- **`ANTHROPIC_API_KEY`** is now required (server `.env`, alongside
+  `HA_WEBHOOK_URL`) — `scraper.get_month_menu()` fails without it, same
+  fail-soft-into-`_status`-error behavior as a Home Assistant push failure.
+  `GET /api/integrations` reports both keys' configured/not-configured
+  state.
 - Settings (calendar pull schedule) persist to `/data/settings.json`, not
-  environment variables — env vars (`CALENDAR_RUN_TIME`, `CALENDAR_RUN_DAYS`)
-  are defaults used only the very first time, before a settings file exists.
-  `load_settings()` merges the file's contents *over* `_DEFAULTS` rather
-  than returning the file verbatim — this matters because the server's
-  settings.json predates this schema (it's from the SMS era) and lacks
-  `calendar_run_time`/`calendar_run_days` entirely; without the merge, the
-  scheduler would silently come up with zero jobs after every deploy until
-  someone happened to re-save the Schedule card in the UI. Verified via a
-  migration test (old-schema settings.json + fresh container → correct
-  defaults filled in, `next_calendar_run` populated).
+  environment variables — env vars (`CALENDAR_RUN_TIME`,
+  `CALENDAR_RUN_DAY_OF_MONTH`) are defaults used only the very first time,
+  before a settings file exists. `load_settings()` merges the file's
+  contents *over* `_DEFAULTS` rather than returning the file verbatim —
+  this matters because the server's settings.json predates several schema
+  changes (SMS era, then the weekly-schedule era) and is missing whatever
+  keys the current schema expects; without the merge, the scheduler would
+  silently come up with zero jobs after every deploy until someone happened
+  to re-save the Schedule card in the UI.
 - **No vacation/no-push date ranges (removed 2026-09-03, briefly existed).**
   A day with no posted menu already produces an empty `items` list from
-  `get_week_menu()`, and `_push_calendar()` already skips pushing when
+  `get_month_menu()`, and `_push_calendar()` already skips pushing when
   `items` is empty — that's sufficient for "don't create an event for a day
   with nothing happening." A separate vacation-range setting to *additionally*
   suppress days the school did post a menu for was judged unnecessary
   complexity and deleted at the user's request; don't re-add it without
   being asked.
 - **Home Assistant integration (built 2026-09-02/03, calendar-only pivot
-  2026-09-03):** `HA_WEBHOOK_URL` is now required, not optional — it's the
-  only way this app does anything externally-visible. On the HA side
+  2026-09-03):** `HA_WEBHOOK_URL` is required — it's the only way this app
+  does anything externally-visible. On the HA side
   (`automation.lunch_menu_family_calendar`, `http://192.168.86.207:8123`),
   a `local_only` webhook trigger runs `calendar.create_event` against
   `calendar.family_2` (the "Family" Google Calendar entity — there are two
@@ -55,38 +84,47 @@ commits), they describe the pre-2026-09-03 app, not the current one.
   `delete_event` — a bad push can't be cleaned up via the HA API, only
   manually in Google Calendar. No duplicate-prevention: re-pushing the
   same date's menu creates a second event — this is why the default
-  schedule is a single day/week, not daily (see below).
+  schedule is a single day/month, not daily (see below).
   **`mode: parallel`, not the default `single` (fixed 2026-09-03 after a
-  real bug):** a week push fires one webhook POST per school day
-  back-to-back with no delay. Under `mode: single`, HA rejects any
-  webhook trigger that arrives while a previous run is still executing
-  (`calendar.create_event` itself takes ~600-700ms) — 4 of every 5 days
-  were silently dropped as `failed_single`, so only one day's event ever
-  actually got created per week push. Found via `ha_get_automation_traces`
-  after the user reported a manual test push not showing up on the
-  calendar. Each invocation is independent (one `calendar.create_event`
-  call, no shared state), so `parallel` is the correct mode per HA's own
-  guidance ("independent per-entity actions → parallel"), not just a
-  workaround. The webhook id is
+  real bug):** a push fires one webhook POST per school day back-to-back
+  with no delay. Under `mode: single`, HA rejects any webhook trigger that
+  arrives while a previous run is still executing (`calendar.create_event`
+  itself takes ~600-700ms) — most days in a batch were silently dropped as
+  `failed_single`, so only one day's event ever actually got created per
+  push. Found via `ha_get_automation_traces` after the user reported a
+  manual test push not showing up on the calendar. Each invocation is
+  independent (one `calendar.create_event` call, no shared state), so
+  `parallel` is the correct mode per HA's own guidance ("independent
+  per-entity actions → parallel"), not just a workaround. The webhook id is
   a random token that lives only in the server's `.env` (`HA_WEBHOOK_URL`)
   and in the automation's trigger config — never committed here. If the
   calendar push stops working, check both ends: `GET /api/integrations`
   on this app reports whether `HA_WEBHOOK_URL` is set at all, and
   `ha_get_automation_traces` on `automation.lunch_menu_family_calendar`
   shows whether HA is even receiving the webhook.
-- **Independent weekly-pull scheduler (2026-09-03):** calendar pulls run on
-  their own schedule (`calendar_run_time`/`calendar_run_days` in settings,
-  `schedule` jobs tagged `"calendar"`), completely decoupled from any other
-  job — there's nothing else in this app anymore, but the tag/isolation
-  pattern (`schedule.clear("calendar")` instead of a bare `schedule.clear()`)
-  is deliberate so a future second schedule wouldn't stomp this one.
-  `push_week_to_calendar()` fetches the whole current school week and
-  pushes every day with a posted menu. Default is Sunday 18:00 (one
-  firing/week) — the UI warns that picking more than one day/week creates
-  duplicate events for days already covered, since there's no dedup.
-- **Manual actions:** "Push This Week to Calendar" (`POST /api/push-calendar`
-  → `push_week_to_calendar()`, same function the schedule calls) and
-  "Preview This Week's Menu" (`GET /api/preview` → `get_week_menu()`
+- **Monthly pull scheduler (weekly → monthly, 2026-09-08):** calendar pulls
+  run on their own schedule (`calendar_run_time`/`calendar_run_day_of_month`
+  in settings, `schedule` jobs tagged `"calendar"`), completely decoupled
+  from anything else — the tag/isolation pattern
+  (`schedule.clear("calendar")` instead of a bare `schedule.clear()`) is
+  deliberate so a future second schedule wouldn't stomp this one. The
+  `schedule` library has no native day-of-month trigger, so
+  `apply_calendar_schedule()` registers a *daily* job
+  (`_maybe_push_month`) that no-ops unless `datetime.now().day` matches
+  the configured day — `push_month_to_calendar()` only actually runs on
+  that one day per month. `next_calendar_run()` computes the real next
+  firing directly from settings (handling month/year rollover and clamping
+  the configured day to short months, e.g. day 30 in February → the 28th)
+  for the UI's "Next Scheduled" field, since the underlying `schedule` job's
+  own `next_run` is meaningless here (it's always just "tomorrow"). Default
+  is day 25 at 18:00 — picked because the district tends to post next
+  month's PDF a few days before month-end, so pulling then usually catches
+  the *upcoming* month rather than the one ending. Whatever month is live
+  at the permalink when the job fires is what gets pushed — there's no
+  logic forcing "next month specifically."
+- **Manual actions:** "Push This Month to Calendar" (`POST /api/push-calendar`
+  → `push_month_to_calendar()`, same function the schedule calls) and
+  "Preview This Month's Menu" (`GET /api/preview` → `get_month_menu()`
   directly, pure dry run, no Home Assistant push, no side effects at all —
   use this to check the scraper before trusting a scheduled push).
 
@@ -97,8 +135,10 @@ pip install -r requirements.txt
 python3 app.py
 ```
 
-Needs `HA_WEBHOOK_URL` set to actually push anywhere; without it the app
-still runs and previews fine, `GET /api/integrations` just reports
+Needs `ANTHROPIC_API_KEY` set for anything menu-related to work at all
+(including Preview, since that calls the same Claude-backed scraper) and
+`HA_WEBHOOK_URL` set to actually push anywhere — without the latter the
+app still runs and previews fine, `GET /api/integrations` just reports
 `ha_webhook_configured: false` and pushes resolve to `"not configured"`.
 Visit `http://localhost:2323`.
 
@@ -124,7 +164,8 @@ manually-edited `.env` on its own.
 **Persistent state, never touched by deploys:** `data/settings.json` (the
 volume-mounted `/data` inside the container) holds the calendar pull
 schedule — this is live user data, not deploy-managed. `.env` holds
-`HA_WEBHOOK_URL` and is server-only, gitignored, never committed.
+`HA_WEBHOOK_URL` and `ANTHROPIC_API_KEY` and is server-only, gitignored,
+never committed.
 
 **Container timezone (fixed 2026-09-08):** the container previously had no
 timezone configured — `python:3.12-slim` defaults to UTC and the

app.py

@@ -2,6 +2,7 @@
 Lunch Notifier — pulls the Mariposa lunch menu and pushes it to a Home
 Assistant calendar. Visit http://localhost:2323 to manage settings.
 """
+import calendar
 import json
 import logging
 import os
@@ -14,7 +15,7 @@ import requests
 import schedule
 from flask import Flask, jsonify, request
 
-from scraper import get_week_menu
+from scraper import get_month_menu
 
 logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
 log = logging.getLogger(__name__)
@@ -27,7 +28,7 @@ PORT = int(os.environ.get("PORT", "2323"))
 # Env-var defaults (used only when no settings file exists yet)
 _DEFAULTS = {
     "calendar_run_time": os.environ.get("CALENDAR_RUN_TIME", "18:00"),
-    "calendar_run_days": [d.strip() for d in os.environ.get("CALENDAR_RUN_DAYS", "sun").split(",") if d.strip()],
+    "calendar_run_day_of_month": int(os.environ.get("CALENDAR_RUN_DAY_OF_MONTH", "25")),
 }
 
 _status = {"last_calendar_run": None, "calendar_result": None}
@@ -41,8 +42,9 @@ def load_settings() -> dict:
         try:
             with open(SETTINGS_FILE) as f:
                 # Merge over _DEFAULTS so a settings.json from before a schema
-                # change (e.g. the pre-calendar SMS-era file) still gets
-                # calendar_run_time/calendar_run_days rather than silently
+                # change (e.g. the pre-calendar SMS-era file, or the
+                # weekly-schedule era before calendar_run_day_of_month
+                # existed) still gets a valid schedule rather than silently
                 # scheduling nothing.
                 return {**_DEFAULTS, **json.load(f)}
         except Exception:
@@ -80,23 +82,23 @@ def _push_calendar(items: list[str], date_str: str | None) -> str:
         return f"error: {e}"
 
 
-def push_week_to_calendar() -> None:
-    """Fetch the current school week and push every day that has a posted
-    menu — days with nothing posted (weekends, breaks, holidays) are
-    skipped automatically since there's nothing to create an event from.
-    Runs on its own schedule (calendar_run_time/calendar_run_days),
-    independent of anything else."""
+def push_month_to_calendar() -> None:
+    """Fetch whichever month is currently published and push every day that
+    has a posted menu — days with nothing posted (weekends, breaks,
+    holidays) are skipped automatically since there's nothing to create an
+    event from. Runs on its own schedule (calendar_run_time/
+    calendar_run_day_of_month), independent of anything else."""
     global _status
 
     try:
-        week = get_week_menu()
+        month = get_month_menu()
     except Exception as e:
-        log.exception("Week fetch failed")
+        log.exception("Month fetch failed")
         _status = {**_status, "last_calendar_run": datetime.now().isoformat(), "calendar_result": f"error: {e}"}
         return
 
     pushed = errors = no_items = 0
-    for day in week:
+    for day in month:
         result = _push_calendar(day["items"], day["date"])
         if result == "ok":
             pushed += 1
@@ -105,7 +107,7 @@ def push_week_to_calendar() -> None:
         else:
             errors += 1
 
-    log.info(f"Week push: {pushed} pushed, {errors} errors, {no_items} no menu")
+    log.info(f"Month push: {pushed} pushed, {errors} errors, {no_items} no menu")
 
     if errors:
         summary = f"{pushed} ok, {errors} error" + ("s" if errors != 1 else "")
@@ -123,26 +125,43 @@ def push_week_to_calendar() -> None:
 
 # ── Scheduler ─────────────────────────────────────────────────────────────────
 
-DAY_MAP_FNS = {
-    "mon": lambda: schedule.every().monday,
-    "tue": lambda: schedule.every().tuesday,
-    "wed": lambda: schedule.every().wednesday,
-    "thu": lambda: schedule.every().thursday,
-    "fri": lambda: schedule.every().friday,
-    "sat": lambda: schedule.every().saturday,
-    "sun": lambda: schedule.every().sunday,
-}
+def _maybe_push_month(day_of_month: int) -> None:
+    """schedule fires this daily at run_time; only actually pushes on the
+    configured day of the month (the `schedule` library has no native
+    day-of-month trigger)."""
+    if datetime.now().day == day_of_month:
+        push_month_to_calendar()
 
 
 def apply_calendar_schedule(settings: dict) -> None:
     with _lock:
         schedule.clear("calendar")
         run_time = settings.get("calendar_run_time", "18:00")
-        for day in settings.get("calendar_run_days", []):
-            day = day.strip().lower()
-            if day in DAY_MAP_FNS:
-                DAY_MAP_FNS[day]().at(run_time).do(push_week_to_calendar).tag("calendar")
-                log.info(f"Scheduled calendar pull: {day} at {run_time}")
+        day_of_month = settings.get("calendar_run_day_of_month", 25)
+        schedule.every().day.at(run_time).do(_maybe_push_month, day_of_month).tag("calendar")
+        log.info(f"Scheduled calendar pull: day {day_of_month} of each month at {run_time}")
+
+
+def next_calendar_run(settings: dict) -> datetime:
+    """The `schedule` job above fires daily (it has no native day-of-month
+    trigger — see _maybe_push_month), so its own next_run is just
+    "tomorrow", not the actual next push. Compute the real one directly
+    from settings for display in the UI."""
+    run_time = settings.get("calendar_run_time", "18:00")
+    hour, minute = (int(p) for p in run_time.split(":"))
+    day_of_month = settings.get("calendar_run_day_of_month", 25)
+
+    now = datetime.now()
+    candidate = now.replace(
+        day=min(day_of_month, calendar.monthrange(now.year, now.month)[1]),
+        hour=hour, minute=minute, second=0, microsecond=0,
+    )
+    if candidate <= now:
+        y, m = (now.year, now.month + 1) if now.month < 12 else (now.year + 1, 1)
+        candidate = candidate.replace(
+            year=y, month=m, day=min(day_of_month, calendar.monthrange(y, m)[1]),
+        )
+    return candidate
 
 
 def _schedule_loop() -> None:
@@ -176,7 +195,7 @@ def api_get_settings():
 def api_post_settings():
     data = request.get_json(force=True)
     s = load_settings()
-    for key in ("calendar_run_time", "calendar_run_days"):
+    for key in ("calendar_run_time", "calendar_run_day_of_month"):
         if key in data:
             s[key] = data[key]
     save_settings(s)
@@ -186,32 +205,33 @@ def api_post_settings():
 
 @app.route("/api/push-calendar", methods=["POST"])
 def api_push_calendar():
-    threading.Thread(target=push_week_to_calendar, daemon=True).start()
+    threading.Thread(target=push_month_to_calendar, daemon=True).start()
     return jsonify({"ok": True})
 
 
 @app.route("/api/preview")
 def api_preview():
-    """Dry-run week scrape: no Home Assistant push, no side effects."""
+    """Dry-run month scrape: no Home Assistant push, no side effects."""
     try:
-        week = get_week_menu()
+        month = get_month_menu()
     except Exception as e:
         return jsonify({"error": str(e)}), 502
-    return jsonify({"days": week})
+    return jsonify({"days": month})
 
 
 @app.route("/api/integrations")
 def api_integrations():
-    return jsonify({"ha_webhook_configured": bool(os.environ.get("HA_WEBHOOK_URL"))})
+    return jsonify({
+        "ha_webhook_configured": bool(os.environ.get("HA_WEBHOOK_URL")),
+        "anthropic_api_key_configured": bool(os.environ.get("ANTHROPIC_API_KEY")),
+    })
 
 
 @app.route("/api/status")
 def api_status():
-    with _lock:
-        next_runs = sorted([j.next_run for j in schedule.get_jobs("calendar") if j.next_run])
     return jsonify({
         **_status,
-        "next_calendar_run": next_runs[0].isoformat() if next_runs else None,
+        "next_calendar_run": next_calendar_run(load_settings()).isoformat(),
     })
 
 
@@ -277,17 +297,11 @@ header p { color: var(--muted); font-size: .875rem; margin-top: 2px; }
 .field:last-child { margin-bottom: 0; }
 label.field-label { display: block; font-size: .8rem; font-weight: 600; color: var(--text);
                     margin-bottom: 6px; }
-input[type=text], input[type=time] {
+input[type=text], input[type=time], input[type=number] {
   width: 100%; padding: 9px 12px; border: 1px solid var(--border); border-radius: 8px;
   font-size: .875rem; color: var(--text); background: var(--card);
 }
 input:focus { outline: 2px solid var(--primary); outline-offset: -1px; border-color: transparent; }
-/* Days toggle */
-.days-row { display: flex; gap: 6px; flex-wrap: wrap; }
-.day-chip { padding: 6px 13px; border: 1.5px solid var(--border); border-radius: 99px;
-            font-size: .8rem; font-weight: 600; cursor: pointer; background: var(--card);
-            transition: all .15s; user-select: none; }
-.day-chip.on { background: var(--primary); color: #fff; border-color: var(--primary); }
 .item-sub { font-size: .75rem; color: var(--muted); margin-top: 1px; }
 /* Divider */
 .divider { border: none; border-top: 1px solid var(--border); margin: 14px 0; }
@@ -297,11 +311,11 @@ input:focus { outline: 2px solid var(--primary); outline-offset: -1px; border-co
          font-size: .875rem; font-weight: 500; opacity: 0; transition: opacity .25s, transform .25s;
          pointer-events: none; white-space: nowrap; z-index: 100; }
 .toast.show { opacity: 1; transform: translateX(-50%) translateY(0); }
-.week-day { padding: 8px 0; border-bottom: 1px solid var(--border); }
-.week-day:last-child { border-bottom: none; }
-.week-day-date { font-size: .8rem; font-weight: 600; margin-bottom: 3px; }
-.week-day-items { font-size: .8rem; color: #c9d1d9; white-space: pre-wrap; }
-.week-day-empty { font-size: .8rem; color: var(--muted); font-style: italic; }
+.day-row { padding: 8px 0; border-bottom: 1px solid var(--border); }
+.day-row:last-child { border-bottom: none; }
+.day-row-date { font-size: .8rem; font-weight: 600; margin-bottom: 3px; }
+.day-row-items { font-size: .8rem; color: #c9d1d9; white-space: pre-wrap; }
+.day-row-empty { font-size: .8rem; color: var(--muted); font-style: italic; }
 </style>
 </head>
 <body>
@@ -330,10 +344,10 @@ input:focus { outline: 2px solid var(--primary); outline-offset: -1px; border-co
   <div class="card">
     <div class="card-label">Manual Actions</div>
     <button id="push-cal-btn" class="btn btn-primary btn-send" onclick="pushCalendar()">
-      Push This Week to Calendar
+      Push This Month to Calendar
     </button>
     <button id="preview-btn" class="btn btn-ghost" style="width:100%;margin-top:10px;" onclick="previewMenu()">
-      Preview This Week's Menu
+      Preview This Month's Menu
     </button>
     <div id="preview-box" style="display:none;margin-top:14px;"></div>
   </div>
@@ -343,8 +357,9 @@ input:focus { outline: 2px solid var(--primary); outline-offset: -1px; border-co
     <div class="card-label">Integrations</div>
     <dl class="stat-grid">
       <div class="stat"><dt>Home Assistant</dt><dd id="i-ha">—</dd></div>
+      <div class="stat"><dt>Anthropic API</dt><dd id="i-anthropic">—</dd></div>
     </dl>
-    <div class="item-sub" style="margin-top:10px;">Configured via <code>.env</code> (<code>HA_WEBHOOK_URL</code>) on the server, not here.</div>
+    <div class="item-sub" style="margin-top:10px;">Configured via <code>.env</code> (<code>HA_WEBHOOK_URL</code>, <code>ANTHROPIC_API_KEY</code>) on the server, not here.</div>
   </div>
 
   <!-- Schedule -->
@@ -355,9 +370,9 @@ input:focus { outline: 2px solid var(--primary); outline-offset: -1px; border-co
       <input type="time" id="f-time" style="max-width:160px">
     </div>
     <div class="field">
-      <label class="field-label">Pull Days</label>
-      <div class="days-row" id="days-row"></div>
-      <div class="item-sub" style="margin-top:8px;">Each day fires a full week pull. Picking more than one day will create duplicate calendar events for days already pushed &mdash; most people want just one (e.g. Sunday evening).</div>
+      <label class="field-label">Pull Day of Month</label>
+      <input type="number" id="f-day" min="1" max="28" style="max-width:100px">
+      <div class="item-sub" style="margin-top:8px;">Fires once a month and pushes whichever month is currently published &mdash; e.g. day 25 catches next month's menu once the district posts it near month-end. There's no dedup, so re-pushing the same month creates duplicate calendar events.</div>
     </div>
     <hr class="divider">
     <button class="btn btn-primary" onclick="saveSchedule()">Save Schedule</button>
@@ -367,8 +382,6 @@ input:focus { outline: 2px solid var(--primary); outline-offset: -1px; border-co
 <div class="toast" id="toast"></div>
 
 <script>
-const DAYS = ['mon','tue','wed','thu','fri','sat','sun'];
-const DAY_LABELS = {mon:'Mon',tue:'Tue',wed:'Wed',thu:'Thu',fri:'Fri',sat:'Sat',sun:'Sun'};
 let cfg = {};
 
 async function load() {
@@ -391,18 +404,16 @@ function fmt(iso) {
 }
 
 function applyIntegrations(ig) {
-  document.getElementById('i-ha').innerHTML = ig.ha_webhook_configured
+  const badge = ok => ok
     ? '<span class="badge badge-green">Configured</span>'
     : '<span class="badge badge-yellow">Not configured</span>';
+  document.getElementById('i-ha').innerHTML = badge(ig.ha_webhook_configured);
+  document.getElementById('i-anthropic').innerHTML = badge(ig.anthropic_api_key_configured);
 }
 
 function applySettings(s) {
   document.getElementById('f-time').value = s.calendar_run_time || '18:00';
-  const row = document.getElementById('days-row');
-  row.innerHTML = DAYS.map(d =>
-    `<span class="day-chip ${(s.calendar_run_days||[]).includes(d)?'on':''}" data-day="${d}"
-           onclick="this.classList.toggle('on')">${DAY_LABELS[d]}</span>`
-  ).join('');
+  document.getElementById('f-day').value = s.calendar_run_day_of_month || 25;
 }
 
 function applyStatus(st) {
@@ -417,8 +428,8 @@ function applyStatus(st) {
 
 async function saveSchedule() {
   const calendar_run_time = document.getElementById('f-time').value;
-  const calendar_run_days = [...document.querySelectorAll('.day-chip.on')].map(c=>c.dataset.day);
-  await post('/api/settings', {calendar_run_time, calendar_run_days});
+  const calendar_run_day_of_month = parseInt(document.getElementById('f-day').value, 10);
+  await post('/api/settings', {calendar_run_time, calendar_run_day_of_month});
   toast('Schedule saved');
   load();
 }
@@ -441,14 +452,14 @@ async function previewMenu() {
     const r = await fetch('/api/preview');
     const data = await r.json();
     if (data.error) {
-      box.innerHTML = `<div class="week-day-empty">Error: ${data.error}</div>`;
+      box.innerHTML = `<div class="day-row-empty">Error: ${data.error}</div>`;
     } else {
       box.innerHTML = (data.days||[]).map(d => `
-        <div class="week-day">
-          <div class="week-day-date">${d.date}</div>
+        <div class="day-row">
+          <div class="day-row-date">${d.date}</div>
           ${d.items.length
-            ? `<div class="week-day-items">${d.items.map(i=>'• '+i).join('\\n')}</div>`
-            : '<div class="week-day-empty">No menu posted</div>'}
+            ? `<div class="day-row-items">${d.items.map(i=>'• '+i).join('\\n')}</div>`
+            : '<div class="day-row-empty">No menu posted</div>'}
         </div>`).join('');
     }
     box.style.display = '';

requirements.txt

@@ -1,3 +1,4 @@
+anthropic==0.125.0
 flask==3.1.0
 requests==2.32.3
 schedule==1.2.2

scraper.py

@@ -1,51 +1,93 @@
+import base64
+import json
 import logging
-from datetime import datetime
 
+import anthropic
 import requests
 
 logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
 log = logging.getLogger(__name__)
 
-SCHOOL_SLUG = "kyrene-de-la-mariposa"
-MENU_TYPE = "lunch"
-API_BASE = "https://kyrene.api.nutrislice.com/menu/api/weeks/school"
+# Stable Finalsite "resource manager" permalink for the K-5 lunch menu, found
+# on https://www.kyrene.org/family-student-resources/food-services/school-menus.
+# This URL itself never changes — it 302-redirects to whichever monthly PDF
+# is currently published (verified 2026-09-08: it pointed straight at that
+# month's file). Nutrislice, the previous data source, was discontinued by
+# the district.
+MENU_PAGE_URL = "https://www.kyrene.org/fs/resource-manager/view/32bb3f8b-5a50-4b38-ad68-a5d3108de571"
+
+MODEL = "claude-opus-5"
+
+_MENU_SCHEMA = {
+    "type": "object",
+    "properties": {
+        "days": {
+            "type": "array",
+            "items": {
+                "type": "object",
+                "properties": {
+                    "date": {"type": "string", "description": "ISO date, YYYY-MM-DD"},
+                    "items": {"type": "array", "items": {"type": "string"}},
+                },
+                "required": ["date", "items"],
+                "additionalProperties": False,
+            },
+        },
+    },
+    "required": ["days"],
+    "additionalProperties": False,
+}
+
+_PROMPT = """This PDF is a school cafeteria's monthly lunch calendar (grades K-5). It's a
+calendar-grid layout with one cell per school day, each listing a Breakfast
+section and a Lunch section.
+
+Extract ONLY the Lunch items (ignore Breakfast) for every day of the month
+shown on the calendar. For each day 1 through the last day of the month:
+- If the calendar shows lunch items for that day, list them.
+- If the day has no cafeteria menu (weekend, holiday, or a no-school day
+  like a break), return an empty items list for that date.
+
+Read the month and year printed on the calendar itself to build the ISO
+dates. Return every calendar day of the month, in date order."""
 
 
-def _fetch_week(date_str: str) -> dict:
-    dt = datetime.strptime(date_str, "%Y-%m-%d")
-    url = f"{API_BASE}/{SCHOOL_SLUG}/menu-type/{MENU_TYPE}/{dt.year}/{dt.month:02d}/{dt.day:02d}/"
-    log.info(f"Fetching {url}")
-
-    resp = requests.get(
-        url,
-        headers={"Accept": "application/json", "User-Agent": "Mozilla/5.0"},
-        timeout=20,
-    )
+def _fetch_current_pdf() -> bytes:
+    resp = requests.get(MENU_PAGE_URL, timeout=20)
     resp.raise_for_status()
-    return resp.json()
+    return resp.content
 
 
-def _extract_items(day: dict) -> list[str]:
-    items = []
-    for entry in day.get("menu_items", []):
-        if entry.get("is_section_title") or entry.get("is_station_header"):
-            continue
-        food = entry.get("food")
-        if food and food.get("name"):
-            items.append(food["name"].strip())
-    return items
+def get_month_menu() -> list[dict]:
+    """Fetches whichever monthly K-5 lunch menu PDF is currently published
+    and asks Claude to read it into structured per-day data. Returns
+    [{"date": "YYYY-MM-DD", "items": [...]}, ...] for every day of that
+    month — days with no posted menu (weekends, breaks, holidays) come back
+    with an empty items list, same contract the old Nutrislice-based scraper
+    used."""
+    log.info(f"Fetching {MENU_PAGE_URL}")
+    pdf_b64 = base64.standard_b64encode(_fetch_current_pdf()).decode("utf-8")
 
+    client = anthropic.Anthropic()
+    response = client.messages.create(
+        model=MODEL,
+        max_tokens=8000,
+        messages=[{
+            "role": "user",
+            "content": [
+                {
+                    "type": "document",
+                    "source": {
+                        "type": "base64",
+                        "media_type": "application/pdf",
+                        "data": pdf_b64,
+                    },
+                },
+                {"type": "text", "text": _PROMPT},
+            ],
+        }],
+        output_config={"format": {"type": "json_schema", "schema": _MENU_SCHEMA}},
+    )
 
-def get_week_menu(date_str: str | None = None) -> list[dict]:
-    """Returns every day in the school-published week containing date_str,
-    as [{"date": "YYYY-MM-DD", "items": [...]}, ...] in the API's own order.
-    Days with no posted menu (weekends, breaks, holidays) have an empty items list."""
-    if date_str is None:
-        date_str = datetime.now().strftime("%Y-%m-%d")
-
-    data = _fetch_week(date_str)
-    return [
-        {"date": day.get("date"), "items": _extract_items(day)}
-        for day in data.get("days", [])
-        if day.get("date")
-    ]
+    text = next(b.text for b in response.content if b.type == "text")
+    return json.loads(text)["days"]