User policy decision (2026-06-05, exact): 'Hard filter. If it is denied
this week twice, it should be considered denied for good.'
The planner had no cross-week memory of denials: a denial on
meal_plan_item.approval_status was never consulted by the planner,
and NeverSuggest (the per-family permanent blocklist) was empty for
the user. The 'Roasted Sweet Potato and Chickpea Bowl' the user
denied on 2026-05-15 was still in the planner's pool 3 weeks
later.
Implements C + Z (explicit two-button model + soft-decay +
hard-filter escalation):
- Approve: untouched.
- Deny this week (1st in 90d): denial_expires_at = now() + 90d.
- Deny this week (2nd in 90d, server-side auto-escalation):
denial_expires_at = NULL + a NeverSuggest row written.
- Never again (explicit): same as the 2nd-time auto-escalation.
Both soft and permanent denials are hard filters in the planner
(per user). A denied recipe never reappears until either the 90d
window expires or the user un-blocks via the NeverSuggest API.
Changes:
- Migration 0016: meal_plan_item.denial_expires_at (partial index)
and meal_plan_vote.denial_scope.
- 3 backend helpers (_apply_denial, _ensure_never_suggest_recipe,
_has_prior_active_soft_denial) — single source of truth for the
deny path.
- POST /api/meals/items/{id}/deny?scope=this_week|never_again
(default this_week). Returns promoted_to_permanent.
- POST /api/meals/vote/{id} extended: vote=approve|deny|never_again.
Returns denial_scope + promoted_to_permanent.
- GET /api/meals/vote/{id} HTML page renders 3 buttons; supports
one-click ?scope=... for email direct-action links.
- Email template (step_email): 3 direct-action links per recipe
plus a secondary 'open vote page' link.
- Planner: _load_blocklists returns 3 sets; soft_denied_recipes
is hard-filtered (union with blocked_recipes at the call site).
- Frontend: MealCard renders 3 buttons (Approve / Deny this week
/ Never again) for pending items. handleDeny is scope-aware;
toast reflects promoted_to_permanent. window.confirm on
'Never again' prevents accidental permanent blocks.
Verification:
- npm run build green.
- 21/21 planner tests pass (1 pre-existing test_filter_blocks_by_cost
failure is NOT introduced by Sprint 8 — verified via git stash).
- Review/sprint8-verification.md: 11-step browser smoke + 4 API
curls + email-render procedure + rollback.
Files:
- backend/alembic/versions/0016_denial_decay_and_scope.py (new)
- backend/app/models/__init__.py:221-242, 250-269
- backend/app/schemas/__init__.py:204-219, 248-269
- backend/app/api/meals.py:30-138 (helpers), 240-330 (HTML page),
380-455 (submit_vote), 486-552 (deny_meal_item)
- backend/app/services/orchestrator/steps.py:283-300
- backend/app/services/planner/generate.py:59-99, 150-194
- frontend/src/api/index.ts:48-58
- frontend/src/pages/Dashboard.tsx:38-50, 385-410
- Review/{sprint8-verification,ui-nielsen-audit,handoff-ui-audit}.md
- fix-ui-audit.md
- docs/HANDOFF.md
- .agent/{plan,context}.md
Deploy (user runs on deployment host):
cd ~/MealPlanner && git pull
docker compose exec backend alembic upgrade head
docker compose -f docker-compose.yml up -d --build backend frontend
User report 2026-06-05: 'webui Meal Planner page is empty' on Friday
morning after the Friday email went out. Root cause: the orchestrator
keyed plans by the most-recent-Friday while the frontend's isoMonday()
returned the most-recent-Monday — a 7-day mismatch on Fridays.
Fixes (one semantic across the stack):
- runner._current_week_start() returns the upcoming Monday (today if
Mon, else the next Mon). The Friday email subject
('Meal plan for week of <date>') automatically picks up the new
value via run.week_start_date.
- frontend isoMonday -> upcomingMonday (same logic; renamed for
intent). isoMonday kept as a deprecated alias.
- New WeekRangeNav component (Dashboard + ShoppingList share it).
Renders [<] Jun 8 - Jun 14 [>] with clickable chevrons and a
clickable range label that jumps to the upcoming week. Replaces
the Sprint 5 inline segmented control on both pages.
- New formatWeekRange(mondayIso) helper (UTC-stable; uses
timeZone: 'UTC' so the rendered date matches the stored ISO date
regardless of viewer TZ; closes a latent bug in formatIsoDate too).
- New SQL fix script that retargets the user's 3-pending-items plan
from 2026-06-05 (Friday-keyed) to 2026-06-08 (upcoming Monday).
Idempotent + transaction-wrapped. Optional block for 2026-05-29.
No backend migration. No new dependencies. Deploy is git pull +
run the SQL fix + docker compose up -d --build backend frontend.
See Review/sprint7-verification.md for the full deploy + smoke flow.
Files:
- backend/app/services/orchestrator/runner.py:20-35
- backend/scripts/fix_2026_06_05_to_2026_06_08.sql (new)
- frontend/src/lib/utils.ts:43-130
- frontend/src/components/WeekRangeNav.tsx (new)
- frontend/src/pages/Dashboard.tsx (3 call sites + 1 segmented control)
- frontend/src/pages/ShoppingList.tsx (5 call sites + 2 segmented controls)
- Review/{sprint7-verification,ui-nielsen-audit,handoff-ui-audit}.md
- fix-ui-audit.md
- docs/HANDOFF.md
- .agent/{plan,context}.md
- Migration 0012 adds score (float) and components (jsonb) to meal_plan_item
- generate.py: populates score and components at create time
- schemas/MealPlanItemResponse: include score + components fields
- GET /api/meal-plans/{id}: returns persisted values instead of zeros
- api/meal_plans.py: /regenerate now passes exclude_recipe_ids into generate_meal_plan
- planner/generate.py: filter recipe_dicts by exclude_recipe_ids set
- image_generation.py: OpenAI gpt-image-1 client with prompt building, b64_json handling
- main.py: StaticFiles mount at /static for generated images
- admin.py: POST /api/admin/trigger-images endpoint for batch generation
- scripts/generate_images.py: CLI for batch image generation
- docker-compose.yml + nginx: volume mounts for static/images persistence
- Verify MealPlanItem.votes ↔ MealPlanVote relationship is correct; no model bug exists
- Add UnitConverter (normalization, within-family, density tables)
- Update cost.py to convert recipe qty to grocery price unit
- Update generate.py _load_match_index to fetch ingredient name + unit
- Fix orchestrator email/shopping-list cost loops to use conversion
- Fix missing Ingredient import in generate.py
- Add 19 unit tests
- config: switch Settings to ConfigDict(extra='ignore') so extra env vars
(spoonacular_api_key, SWIFTLY_BEARER_TOKEN) don't crash import.
Remove deprecated class Config.
- email: wrap SendGrid imports in try/except so the module loads without
the optional dependency. Update test_email_backend to patch Mail/RepyTo.
- planner_select: default PlannerConfig.set_size=21 (3 meals/day × 7) is
way too large for the unit test assertion that checks 3-recipe diversity.
Introduced _CFG_3 with set_size=3 and applied to all tests.
- Delete stale test_matcher.py importing removed functions.
Full suite: 46 passed, 74 skipped (Postgres), 0 failed, 120 collected.
- /api/admin/test-email now calls get_email_backend().send() instead of only logging.
- /api/meals/vote/{id} GET now queries MealPlanVote and renders 'already voted' confirmation if found.
- api/meals.py: fix remaining uppercase MealPlanItemStatus enum ref (DENIED, APPROVED, PENDING).
- Fixes the 'all meals show as pending' status regression and the 'Error: Already voted' bug.
Matcher improvements (matcher.py):
- Plural normalization: 'tortillas'→'tortilla', 'thighs'→'thigh' so
subset recall check works without stemmer
- Precision floor lowered 0.45→0.30: allows 'Bacon'→'Wright Brand Bacon'
(1/3=0.33) while exclusion words still block category contaminants
- _EXCLUSION_WORDS now normalized through same singularizer for consistency
LLM second-pass (llm_matcher.py):
- run_llm_match_job(): for each still-unmatched ingredient, collects top-12
candidates from grocery catalog ranked by fuzzy×precision (same metric as
AUTO matcher), then asks Ollama to pick the best match
- Candidate scoring: combined = (partial_token_sort_ratio/100) × precision
ensures "McCormick Black Pepper" outranks "Dr Pepper" for 'Black Pepper'
- Stores picks as source='auto_llm' (confidence=0.750)
- Ollama Cloud endpoint: https://ollama.com/v1, model: kimi-k2.6:cloud
Migration 0010: adds 'auto_llm' to ingredient_match_source_enum
Config: OLLAMA_BASE_URL / OLLAMA_API_KEY / OLLAMA_MODEL settings
Docker-compose: wires all three Ollama + Spoonacular env vars to backend/scheduler
Scraper service: calls run_llm_match_job after run_match_job on every scrape
Results: AUTO matcher went from 36→25 unmatched (plural normalization fix),
LLM added 3 more (Black Pepper, Zucchini, Chicken Thighs).
Remaining 22 are genuine Lucky CA catalog gaps (standalone olive oil,
dried spices, etc. not in Swiftly weekly ad).
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
Scraper: remove price guard in map_product so produce items without a
catalog price (e.g. Fresh Garlic, Lime sold by weight) are saved to
grocery_item with current_price=NULL rather than skipped.
Matcher:
- Add exact-name fast path: build a lowercase-trimmed name→index map
and skip fuzzy search entirely when the ingredient name matches a
grocery item exactly. Lime → Lime (confidence 1.0), Garlic → Fresh
Garlic from fuzzy (confidence 1.0).
- Add exclusion words: juice, gelatin to prevent beverage/dessert
products from matching cooking ingredients.
- Increase fuzzy candidate limit 20→100 so exact-name items buried in
large tie groups are not missed.
- Add 'juice' to exclusion: prevents '100% Lime Juice' from winning
over plain 'Lime'.
Result: all recipe ingredients now match correct Lucky CA products or
show '—' (no match); zero category cross-contamination remaining.
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
- Add exclusion words: soda, rotisserie, tuna/tonno/salmon/sardine/anchovy
to prevent beverages, prepared poultry, and seafood-in-oil from matching
raw cooking ingredients
- Add min precision floor (0.45): grocery sig-word count must be ≤ 2× the
ingredient's sig-word count, catching long branded products that pass
the word-overlap recall check but are clearly wrong category matches
(e.g. "Garlic Herb Rotisserie Chicken" precision=0.25 now rejected)
Result: all previously wrong matches now show '—' (no match) rather than
a wrong product; correct matches unchanged
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
- Flip matching direction: iterate ingredients, search grocery items
(previously: iterate grocery items → false positives from partial word
overlap, e.g. Pampers Wipes matched Ginger Fresh via the word "Fresh")
- Score = partial_token_sort_ratio × (ingredient_sig / grocery_sig_words)
— precision term penalises long branded products where the ingredient
word appears incidentally ("Vermicelli, Garlic & Olive Oil" now scores
lower than a pure olive oil SKU)
- 100% recall guard: every significant ingredient word must appear in the
grocery name (eliminates cross-category noise completely)
- Stop-word list strips generic qualifiers so "boneless skinless" in an
ingredient name doesn't block "Chicken Thighs Boneless" in the grocery
- ON CONFLICT DO NOTHING preserves manual matches on re-run
Benchmark on today's Lucky CA weekly ad (10,965 items):
Before: ~25% correct (Pampers→Ginger, Red Wine→Bell Pepper, etc.)
After: ~80% correct; remaining misses are data gaps (Lucky has no
standalone garlic or olive oil in this week's ad)
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
- Vote email: recipe cards now include a collapsible <details> block with
numbered cooking steps (recipe.instructions ARRAY)
- Shopping list email: ingredients now grouped under each meal heading
instead of a flat deduplicated list
- step_finalize: fix grocery price lookup (.price -> .current_price)
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
Preloads Ingredient names from the DB (ingredients JSONB has no name field),
deduplicates by ingredient name, looks up top-confidence IngredientGroceryMatch
per ingredient, and renders a rich 4-column HTML table (Ingredient | Qty | Unit |
At Lucky | Price) with an estimated total.
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
recipe.ingredients JSONB has ingredient_id but no name field; preload
names in a single bulk query before the per-member loop so ing_rows,
shopping preview, and cost lookup all render real ingredient names.
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
Wraps the SessionLocal body in try/finally so db.close() is always
called, preventing connection leaks on exception. Updates the
test monkeypatch to use a no-op close() proxy so the transactional
fixture stays live after run_step returns.
Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
Replaces the static SWIFTLY_BEARER_TOKEN env-var lookup with a JIT
mint via the Firebase Identity Toolkit signUp endpoint, gated by the
firebaseApiKey published in luckysupermarkets.com/config.json.
- get_token(): returns cached JWT if exp > now+300s, else mints
- mint_anonymous_token(): fetches API key, posts signUp with
Origin/Referer headers, validates iss + exp on the returned JWT
- SwiftlyAuthMintError surfaces verbatim to ScrapeLog.error_message
- Process-local cache only; threading.Lock around mutate
Tests: 4 unit tests covering fresh mint, cache hit, near-expiry
re-mint, and Firebase non-200. Full suite: 92/92 green.
Spec: docs/specs/2026-05-06-swiftly-token-auto-mint.md
Wiring into lucky_ca_scraper deferred to AM-2.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
MealPlanner
admin