name: q1-2026-operator-checklist type: brand-asset / operator-handoff status: drafted 2026-05-12 — consolidates every action item across the 18-commit Q1 pass-4 session into a single owner-grouped punch-list. Each item names the artifact, the effort estimate, and what it's blocked on. Print this; check items off as they ship. references: every doc/* + migration produced this session — see "Source artifacts" map at the bottom (§10).

Operator action checklist — Q1 → Q4 2026

The 18 commits this session produced 8 standalone docs + 3 migrations + 1 strategy-doc pass + 2 vault placements. This checklist makes that work executable — one owner per item, one box per action, one dependency tag per blocker.

0. The critical-path dependency

Almost every Q1 P0 item is one of three states:

• Independent — ships today. No external blocker.
• Blocked-on-philosophy-name — waits for Michelle's Yue-do final approval. One decision now unlocks 9 surfaces (was 7 — the §1 [DISH-DEFENSE] + [STRUCTURAL-DEFENSE] sentence-items added 2026-05-13 take the cascade-count to 9).
• Blocked-on-supabase-db-push — waits for operator authorization to apply pending migrations. Carlos / engineering moves once approved.

The single highest-leverage move: Michelle reads docs/yue-do.md and confirms Yue-do (or picks an alternative from the brainstorm). 90-min decision. Pairs naturally with the two Saturday defense-sentence items (§1) which cost ~20 min on top.

1. Michelle — owner decisions (single critical-path)

The most cascading owner decisions. Most other items wait on these.

• Approve Yue-do (粵道) as the canonical philosophy-name. APPROVED by Michelle 2026-05-13 — Yue-do (粵道) confirmed; brainstorm alternatives (Yue-wa / Wok-counter / Maasstraat-omakase) closed. Migration 0072_brand_audiences_philosophy_name_seed.sql written + applied to live Supabase; all 6 brand_audiences rows now carry philosophy_name='Yue-do' + per-tier rationale. Julius warn-level enforcement active. Cascade now unblocked: §4.6 P0 #1 (title-tag rewrite) + #6 (LinkedIn company page) + #7 (about-us body rewrite) + §4.3 row 5 essay (position essay) + §5.0.1 rule 1 (cross-channel deployment) + the Grimm Mid-Autumn pitch's one-link rule + the Broekaert position-claim pitch + cast one-pager status flip. Wave 2 deployment can begin per §11.
• Sourcing essay fill-in session with Penn (Saturday morning, ~3 hrs). Fill in docs/sourcing-essay-mar-2026.md §3 checklist: named suppliers, dated visits, dish-supplier pairings, one named refusal anecdote. Unblocks: Versprille press pitch, sustainability_nl awards eligibility, Tactic 1.2 supplier carousel, brand_dishes.ingredients cross-link. Effort: 3 hrs + Penn voice-pass.
• Position essay fill-in session with Penn (Saturday morning, ~4-5 hrs). Fill in docs/position-essay-sep-2026.md §3 checklist: Michelle's Cantonese lineage, Japanese-technique training context, three Yue-do dishes for §G integrity check. Unblocks: Broekaert NRC press pitch, the canonical philosophy-defense LLM-grounding chunk. Effort: 4-5 hrs (longer than sourcing — philosophical extraction is harder).
• Cast roster confirmation — 8 of 9 §7.1 rows in the strategy doc need named individuals (sommelier, pastry lead, sous A/B, FOH lead, bar lead, pastry assistant, kitchen porter). Resolves naturally in Sun's cast 1:1s. Without this, §7 is aspiration; with it, §7 ships as operational SOP. Effort: Sun-led; Michelle confirms names.
• Mid-Autumn 2026 dish identity — the Grimm press pitch references "salt-baked-egg-yolk mooncake with koji glaze" as a placeholder. Confirm Michelle's actual signature mooncake for the Sept 25 pop-up. Penn cannot pitch a dish that doesn't exist. Effort: 1 conversation.
• Mid-Autumn dish-defense sentence in Michelle's own voice — the Grimm pitch §2.8 placeholder [DISH-DEFENSE]: 2-3 sentences naming what the Japanese gesture does to the Cantonese base, no "fusion" framing. Penn's working default exists but Versprille's "authenticity is slippery" warning demands Michelle's own words. Pairs with the dish-identity item above; same conversation. Effort: same conversation, ~10 min on top.
• Cuisine-order structural-defense sentence in Michelle's own voice — the Broekaert pitch §3.8 placeholder [STRUCTURAL-DEFENSE]: 2-3 sentences naming why Cantonese stays primary and Japanese is grammar-not-equal-partner (§1.7 cuisine-order rule materialized). Penn's working default exists but Broekaert reviews the position-claim itself — the words have to be Michelle's because he'll pressure-test them at the table and the email defense must match the table defense. Different sentence than the Mid-Autumn dish-defense; pairs naturally in the same conversation but answers a different question. Effort: ~10 min in the same Saturday session.
• Signature dish identity — per Tactic 4.3 + §4.3 row 4 + §3.3 Mott 32 pattern. Name the one Michiu dish that is the answer to "what is Michiu." Feeds the Jul 2026 essay + every cardinal surface. Effort: 1 conversation.
• 2012 founding-year continuity — Creditsafe + businessrestaurants.nl show Restaurant Michiu B.V. incorporated 2012. Confirm continuous operation at Maasstraat 102 since then. If yes, every long-form essay can use "14 jaar" as a heritage anchor. §3.2 Pak Loh future-reference benefits; press pitches use this. Effort: yes/no answer.
• €25/p package fate — the businessrestaurants.nl directory listing shows morning/afternoon/evening packages at €25/p. Keep / retire / reposition. Penn cannot pitch the §0 shift #2 corporate-dining premium narrative cleanly while a €25/p offering is the directory-visible entry. Effort: 1 decision.
• FOH lead identity — for the Quintonil dual-brand pattern (§7.0.1 + §5.0.2 + §5.0.1 rule 3 + position essay §D). Naming the FOH lead + role title unlocks LinkedIn cast-feature cadence + the position essay's cast-distribution defense. Effort: 1 conversation.
• Dutch LinkedIn profile claim — confirm whether michelle-zhang-07a9ba4 is your personal LinkedIn or share the actual URL. Penn cannot cadence essays on Michelle's personal page without the URL. Effort: 5 min.
• Cultural-moments TBD resolutions — review the 3 proposed TBD resolutions in docs/cultural-moments-seed-2026.md §2.6 (Dragon Boat 2026 → skip; Momiji 2026 → participate-primary; Oseibo 2026 → skip). Approve or override. Effort: 15 min.
• GaultMillau 2027 cast submissions — per 0062_awards_pipeline.sql seeded gault_millau_nl track + the cultural-moments seed §2.4 GaultMillau 2027 row: submit chef + host + sommelier for individual recognitions. Submission window verify Q4 2026. Each submission requires Michelle's per-track approval per §5.5 + §8.4. Effort: per-track decision in Q4 2026.
• Sake program structure decision (NEW 2026-05-13) — docs/sake-essay-may-2026.md §3.1 flags 3 unresolved unknowns gating the entire essay: (a) does a 12-week structured program currently exist OR will it be launched? (b) if launching, when does cycle 1 start? (c) what's the engagement-path price-structure (avond / blok / hele traject)? Penn cannot draft until these resolve; without resolution, the May 2026 publication slot slips to June. Owner-decision = Michelle + named sommelier together (downstream of cast roster confirmation). Effort: 1 hr conversation between Michelle + sommelier once sommelier identity is named.
• Leveranciersnota card launch decision (NEW 2026-05-13) — docs/signature-dish-essay-jul-2026.md §F proposes a per-signature-dish supplier-card pattern (Atomix §3.3 #1 extended): every guest who orders the signature dish gets a small card listing the named suppliers + locations + zaterdag-ochtenden. Yes/no decision; if yes, ~12-week lead-time to design + print + roll out. Owner-decision = Michelle + Anna (FOH-side ergonomics matter). Pairs with the signature-dish identity decision in the same Saturday session. Effort: 1 decision (~5 min on top of the dish-identity discussion); commission timeline if yes.

2. Carlos — site / Schema / GBP / engineering

Ordered by dependency: independent first, then philosophy-name-gated, then post-migration items.

2.1 Independent (ship today)

• Cloudflare crawler allowlist for michiu.nl (§4.6 P0 #5) — twice-re-verified 503; binding constraint on the entire §4.3 essay investment. Pass-4 session (2026-05-11) got HTTP 503 on every WebFetch attempt to michiu.nl/ and michiu.nl/en/about-us/. Re-verified 2026-05-12: both URLs still return 503. The Cloudflare bot-rule is gating anything non-browser-shaped, which means Googlebot / OAI-SearchBot / PerplexityBot / ClaudeBot / GoogleOther-AI are hitting the same wall — every Yue-do essay we publish is invisible to the LLM answer-set until this fix lands. Verify in Cloudflare bot-management dashboard; allowlist Googlebot, OAI-SearchBot, PerplexityBot, ClaudeBot, GoogleOther-AI. Post-fix verification (Carlos runs from a shell to self-confirm before marking ✅): curl -A "Googlebot" -I https://michiu.nl/ returns HTTP/2 200 (repeat with each UA above + https://michiu.nl/en/about-us/). Anything 403/503 means the rule didn't catch that UA. Effort: 1-2 hrs.
• GBP primary-category audit + correction (§4.6 P0 #4). Currently presumed wrong (Yelp says "Thai"; expect similar GBP miscategorization). Change to "Cantonese restaurant" with secondary "Japanese restaurant" — Cantonese first per §1.7. Needs GBP login. Effort: 1 hr.
• hreflang tags verified on michiu.nl — NL primary, EN secondary, both pointing at correct URL paths. Effort: 30 min.
• Schema.org Restaurant JSON-LD audit — verify current state of structured data on michiu.nl/ and /en/home-en/. (Will be replaced by the full markup in docs/michiu-nl-rewrites-q1-2026.md §7 once philosophy-name is approved.) Effort: 30 min.

2.2 Blocked on philosophy-name approval (ship after Michelle confirms Yue-do)

• Title-tag rewrite (§4.6 P0 #1). Current: MICHIU Lounge • Culinary Asian Cuisine - Amsterdam. Target: Michiu · Yue-do · Cantonese-Japanese · Amsterdam-Zuid (52 chars). Effort: < 1 hr.
• <meta description> rewrite (NL + EN per docs/michiu-nl-rewrites-q1-2026.md §2.2 + §2.3). Effort: < 1 hr.
• About-us body rewrite (§4.6 P0 #7) — Penn drafts (full draft in docs/michiu-nl-rewrites-q1-2026.md §3); Michelle approves; Carlos ships to michiu.nl/en/about-us/ (NL + EN). Effort: < 1 day after approval.
• GBP description NL primary + EN secondary — full rewrite with Yue-do in opening sentence (per docs/yue-do.md §3 surface table). Effort: 1 hr.
• GBP CN + JP descriptions — minimal; include 粵道 in characters + target-language gloss per docs/yue-do.md §6.4. Effort: 1 hr.
• Schema.org Restaurant JSON-LD full update per docs/michiu-nl-rewrites-q1-2026.md §7.1 — including alternateName: "Yue-do" so LLM crawlers index the philosophy-name as a primary lookup token. Effort: 1 hr.

2.3 Blocked on supabase db push authorization

• supabase db push for pending migrations: 0053 (Campaigns) + 0058 (lifecycle_outreach_log) + 0060 (philosophy_name) + 0061 (brand_visual_identity) + 0062 (awards_pipeline) + 0070 (cultural_moments seed extension — pending Michelle/Sun TBD review per §1). Effort: 30 min once authorized. 0071 was previously listed here for a brand_dishes.is_signature_flag column — resolved 2026-05-13 via the 0034 schema reconciliation: the column already exists as is_signature BOOLEAN (shipped in 0034_brand_dishes.sql); 8 dishes are already seeded with is_signature=true; no new migration needed.
• brand_dishes.ingredients cross-link populated once sourcing essay v1 ships. Each named dish in essay §C-G → corresponding brand_dishes row → ingredients JSONB with supplier + relationship-since fields. Effort: 1-2 hrs once essay live.

3. Penn — long-form + press + LinkedIn

3.1 Independent

• Create the Michiu LinkedIn company page (§4.6 P0 #6). Full launch copy in docs/michiu-nl-rewrites-q1-2026.md §4 (tagline + NL/EN about). Logo from brand-assets/01-logos/; cover image natural-light kitchen photo from brand-assets/04-photography/staff/ (until brand_visual_identity.status='active'). Effort: 1-2 hrs.
• Confirm Michelle's personal LinkedIn URL (mirrors Michelle's §1 owner-decision item — Penn drives the asking). Effort: 5 min.

3.2 Sourcing essay arc (Mar 2026 v1 + autumn v2)

• Schedule + run Michelle's 3-hr Saturday-morning fill-in session for docs/sourcing-essay-mar-2026.md §A/§C/§D/§E/§F/§G placeholders. Effort: 3 hrs Penn + 3 hrs Michelle.
• Voice-pass the filled-in draft + Julius banned-phrase check + Michelle final approval per §8.4 escalation. Effort: 4 hrs.
• Hand to Carlos for publication at michiu.nl/posts/sourcing. Effort: Carlos task.
• Versprille pitch (Volkskrant) — Week of 8 Sep 2026 (Tue 8 Sep 09:30 primary; Wed 9 Sep 10:00 backup per docs/penn-press-pitches-q1-2026.md §4.12). Pitch is now fully drafted (§4.8 NL + §4.9 EN parallel + §4.10 7-component dossier-anchored pre-read packet + §4.11 visit logistics + §4.13 pre-flight checklist + §4.14 follow-up cadence + §4.15 outcome scenarios + §4.16 owner-input punch-list). Three binding prerequisites: (a) sourcing essay v1 live on michiu.nl/posts/sourcing (overdue per the items above); (b) v2 autumn update drafted to embargo-ready state (Saturday session below); (c) supplier dossier compiled per §7. Without all three, pitch cannot ship. Effort: 1 hr to send once prereqs land.
• Autumn 2026 v2 update — week of 22 Sep 2026 publication target; outline drafted at docs/sourcing-essay-sep-2026-v2.md (2026-05-12). Table-centric structure: §C dossier is the centerpiece (8-18 rows × 7 columns: Naam / Plaats / Sinds / Wat / Cadans / Gerechten / Notities). Mandatory §4.1 compatibility check with v1 before ship — every v1-named supplier reconciled against v2 §C. §E producer-receives-100% accounting is conditional on Michelle adopting the Narisawa zero-mediator pattern by autumn 2026; drops from v2 if not adopted. Saturday session is 2 hrs (shorter than v1's 3 hrs because framing is established). Effort: 2 hrs Penn + 2 hrs Michelle + 1 hr Anna for the dossier rows; voice-pass + Julius + Michelle final = 4 hrs Penn.

3.3 Position essay arc (Sep 15 2026 publication)

• Schedule + run Michelle's 4-5 hr Saturday-morning fill-in session for docs/position-essay-sep-2026.md §C/§D/§E/§G placeholders. Effort: 4-5 hrs each side.
• Voice-pass + Julius check + Michelle approval. Effort: 6 hrs.
• Hand to Carlos for publication at michiu.nl/posts/yue-do — target 15 Sep 2026. Effort: Carlos task.
• Broekaert pitch (NRC) — Week of 10 Aug 2026 (per docs/penn-press-pitches-q1-2026.md §3). Subject line + opening paragraph drafted with the structural-claim framing. Effort: 1 hr.

3.4 Grimm pitch (Het Parool) — earliest of the three

• Grimm Mid-Autumn pop-up pitch — Week of 17 Jun 2026 (per docs/penn-press-pitches-q1-2026.md §5). Subject line + opening paragraph drafted. Requires Michelle's Mid-Autumn dish identity confirmation first. Effort: 1 hr after dish confirmed.

3.5 Other ongoing

• EN translations of the NL-only site copy that doesn't yet have full pairs (IG bio, GBP description, businessrestaurants.nl listing — about-us and LinkedIn already have EN drafts in docs/michiu-nl-rewrites-q1-2026.md). Effort: 2 hrs. Resolved 2026-05-13 via adfa3a2: title-tag NL parallel + LinkedIn tagline NL parallel + businessrestaurants.nl cuisine-field NL primary added; IG bio + TikTok bio EN press-kit references added. Deployment-EN gaps are now zero. Only remaining gap is GBP description — no draft of any language exists in michiu-nl-rewrites-q1-2026.md; deferred to Carlos §2.2 GBP description rewrite once philosophy-name approved (Penn or Carlos drafts at that point).
• Update businessrestaurants.nl listing copy (§5.5 Q1 P0) — full draft in docs/michiu-nl-rewrites-q1-2026.md §5. Effort: 30 min once philosophy-name approved.
• Awards-pipeline submission packets per 0062_awards_pipeline.sql seeded rows — each requires Michelle's per-submission approval (§5.5 + §8.4). 2027 cycle prep starts Q4 2026. Effort: per-submission, 4-6 hrs each.
• Group-booking confirmation template — booker-quote consent line (NEW 2026-05-13) — docs/rai-corporate-essay-nov-2026.md §8 flags this as immediate Penn action. Add a soft-consent line to the standard group-booking confirmation email: "we may share an anonymous summary of your evening in our year-end retrospective." Without consent collected at booking time, the Nov 2026 §F event-recap paragraphs cannot quote booker reactions. Required for Q3-Q4 2026 bookings as they happen — not retroactively obtainable. Effort: 30 min template update + 1 send to Anna for FOH-side awareness.

4. Zuck — social

4.1 Independent (ship today)

• IG bio location fix (§4.6 P0 #2). Current: "Amsterdam Centrum"; target: "Amsterdam-Zuid · Maasstraat 102 · 5 min van RAI" (without philosophy-name yet). Effort: < 10 min.
• TikTok bio location fix same as IG. Effort: < 10 min.

4.2 Blocked on philosophy-name approval

• IG bio + TikTok bio philosophy-name update — full text in docs/michiu-nl-rewrites-q1-2026.md §6.2 / §6.3. Effort: 10 min each.

4.3 Ongoing content discipline

• Retire AI-slop TikTok content — audit existing @michiu_nl TikTok posts per §5.0.2 finding (the "Did you ever try Toro?🤤🤤🤤🤤" pattern). Remove or archive worst offenders before launching new cadence per §5.2. Effort: 1-2 hrs audit.
• Cross-post sourcing essay as IG carousel + LinkedIn long-post once v1 ships (per §5.5 essay_excerpt_linkedin Power-Up). Effort: 1 hr per essay.
• Cross-post position essay similarly post-Sep 15 publication. Effort: 1 hr.
• Visual-identity vocabulary commission — once Sun + Michelle decide on direction per §6.8.2; Zuck coordinates with the commissioned artist + tracks brand_visual_identity.surface_coverage. Effort: 12-16 weeks from commission to first surface live.

5. Sun — strategy / cultural moments / Campaigns

• Cast 1:1s for §7.0.1 cast-roster confirmation + §11.1 hypothesis #6 (cast TikTok willingness). Question set + facilitation guide + session-log template + 9-row sequencing live at docs/cast-1on1-question-set.md. Effort: ~9 hrs total spread across 5 weeks (60-min slot per cast member × 9 cast members; 1-2 sessions/week cadence).
• Cultural-moments TBD review — confirm or override the 3 proposed resolutions in docs/cultural-moments-seed-2026.md §2.6. Effort: 30 min.
• Verify approximate RAI event dates flagged in docs/cultural-moments-seed-2026.md §2.1 against rai.nl/en/calendar (KubeCon, HR Tech, AI/BD Expo, IBC explicitly approximate). Effort: 30 min.
• First Yue-do-anchored Campaign brief — once philosophy-name approved + migration 0053 (Campaigns) is applied live, Sun creates the first Campaign archetype from the §6.5 list (recommend lunch-dice-2026 as the lowest-risk first). Protocol playbook live at docs/lunch-dice-protocol.md (2026-05-12) — covers mechanic design, trigger/location/initiator rules, §4 reward strawman (Michelle approves), §5 FOH service language (Anna voice-passes; banned-framings explicit per gamification-felt-not-told pillar), §6 Circle × dice double-felt-not-told handling, §7 edge cases (refusal / multi-roll / press visitors), §8 capture loop (no PII; no revenue-attribution measurement; that would break the pillar), §9 4-week lunch-only pilot plan + Phase 2 retire option, §10 7-item owner-input gating list, §11 8 structural anti-patterns. Effort: 2 hrs Sun for Campaign archetype after Michelle's §10.1 reward-design decision + Anna's §10.4 pilot-launch decision.
• Visual-identity direction — Sun creates the first brand_visual_identity row (status='planning', vocabulary_summary describing intended direction) once migration 0061 is applied. Pre-commission step. Effort: 1-2 hrs.
• Brief Penn for Mid-Autumn pop-up pitch — Week of 17 Jun 2026 lead-time per docs/penn-press-pitches-q1-2026.md §5. Effort: 30 min.
• Brief Penn for Broekaert position-essay pitch — Week of 10 Aug 2026 lead-time. Effort: 30 min.

6. Siren — review-management

• Yelp category fix (§4.6 P0 #3) — "Thai" → "Cantonese, Japanese, Asian Fusion". Effort: < 30 min.
• TheFork NL / Eet.nu / Iens review-response management — assigned to Siren per §4.2; ongoing. Effort: 1-2 hrs/week ongoing.
• Add Yue-do mention pattern to review-reply prompt — only when reviewer mentions cuisine framing (per docs/yue-do.md §6.6). Update src/brand/review-prompt.ts post-approval; bump PROMPT_VERSION. Effort: 1 hr. Shipped 2026-05-13: added YUE_DO_CONDITIONAL section to src/brand/review-prompt.ts between RETURNING_GUEST and OUTPUT_FORMAT; PROMPT_VERSION bumped v2.0 → v2.1; file docstring updated with v2.1 entry. Captures the §6.6 felt-not-told discipline (default: omit Yue-do; conditional: when reviewer opens the door with cuisine-framing language, naturally include it). Worked example phrasings provided in both EN and NL. pnpm typecheck clean; pnpm regress ran 40/42 structural checks passing (1 voice-mismatch failure on case 7 was unrelated to Yue-do — voice-selection logic untouched; Anthropic API credits low so most cases ran on mock fallback, output comparisons not meaningful). Next review reply generated will tag PROMPT_VERSION='v2.1'.

7. Anna — FOH + DM

• Supplier dossier compilation for the Versprille pitch artifact + v2 sourcing essay §C table — 8-18 structural suppliers minimum, NL-located, dated relationships, named visit dates 2026. Structured capture template + 4-phase workflow at docs/supplier-dossier-template.md (Carlos POS pull → Michelle Saturday interview → supplier-direct consent contact rolling Jul-Aug → press-readiness sort). Press-flag scheme (green / yellow / red / pending) for partial-consent handling. Effort revised: ~10 hrs Anna total over 8 weeks + 2 hrs Carlos (Phase 1) + 2 hrs Michelle (Phase 2 Saturday) + 2-3 hrs Penn voice-pass for §6 expanded packet narratives. Original "2-3 hrs" estimate omitted the Phase 3 supplier-confirmation phase which is the longest-duration phase (4 weeks elapsed time) because suppliers respond on their own clocks. Phase 3 must start no later than mid-July 2026 to leave buffer for non-responders before the 22 Sep v2 publication target.
• FOH playbook documentation as the §6.4 awards_nomination_calendar eligibility surface for the Art of Hospitality award track. Per 0062_awards_pipeline.sql notes — needs ≥2 full operating years of documented Circle gestures + window-seating discipline + service protocols before Penn flags ready-to-submit. Effort: ongoing.
• DM bio update with Yue-do once philosophy-name approved + per docs/yue-do.md §6.5 Anna guidance (phrase in email signature; not lead in conversational DM). Shipped 2026-05-16: DM_YUE_DO_CONDITIONAL added to src/brand/dm-prompt.ts; bumped dm-v2 → dm-v3. Conditional trigger only when guest explicitly asks about cuisine framing; default omit.

8. Clara — lifecycle

• Email signature update with Yue-do per docs/yue-do.md §3 surface table. Shipped 2026-05-16: sign-off in src/lib/marketing-os/lifecycle/email-prompt.ts extended to "het Michiu team · Yue-do · Maasstraat 102"; thanks-v2 → thanks-v3. Email channel only; WhatsApp sign-off unchanged per §6.7.
• Birthday flow updated with Yue-do mention pattern (in signature, not in body). Shipped 2026-05-16: same email-prompt.ts sign-off change applies to all lifecycle templates using the shared builder (post-visit thank-you, birthday flow, lapsed-winback, vip-milestone, review-re-engagement). Body copy untouched.
• LNY 2027 priority send — first Clara campaign once customers projection (OS-22b1 ✅) + lifecycle_outreach_log (OS-22b2 ✅) + Resend integration are all live. Targets Circle members 72h before public LNY long-table release. Effort: 2-3 hrs.

9. Insight — measurement

• Run §4.5 quarterly GEO test query suite Q2 2026 — compare against the Q1 baseline (0.00 across 9 queries measured pass-4 May 2026). Aim for first non-zero result on at least 3 queries by end of Q2. Effort: 1-2 hrs per quarter.
• Philosophy-name deployment coverage tracking — % of cardinal surfaces (per docs/yue-do.md §3) using the canonical phrase, audited monthly. New §10.2 leading indicator. Effort: 30 min monthly.
• Visual-identity surface coverage tracking — once brand_visual_identity.status='active', monthly audit of surface_coverage JSONB. New §10.2 leading indicator. Effort: 30 min monthly post-active.
• Michelle-fronted cadence ratio tracking — trip-wire at ≥80% per §5.0.1 rule 4 / §10.2. Weekly Insight digest carries this on the Health row of the Michelle dashboard widget. Effort: built into Power-Up registry.
• §10.4 Michelle Monday-morning dashboard widget — 4 rows (SB lane + CC lane + CB lane + Health row). Insight builds it. Effort: engineering work; not in this session's scope.

10. Source artifacts map

Every checkbox above maps to one or more source artifacts from this session's 18 commits. If you need to read the long version of any item:

11. Sequencing — recommended order of operations

Today (Wave 1 — independent items):

• Michelle reads docs/yue-do.md (90 min)
• Zuck fixes IG + TikTok bio location (20 min)
• Siren fixes Yelp category (30 min)
• Carlos fixes GBP category + verifies hreflang + audits Schema.org (3 hrs)
• Carlos sets up Cloudflare crawler allowlist (1-2 hrs)
• Penn creates the LinkedIn company page (1-2 hrs)
• Anna kicks off supplier-dossier compilation (Phase 1 Carlos POS pull, ~2 hrs); full effort revised to ~10 hrs over 8 weeks per §7.1 + docs/supplier-dossier-template.md

This week (Wave 2 — philosophy-name-gated, assumes Michelle approves Wave 1):

• Penn drafts title tag + meta description + about-us body rewrite (4 hrs)
• Michelle approves rewrites (1 hr)
• Carlos ships rewrites + Schema.org full update (1 day)
• Zuck updates IG + TikTok bios with philosophy-name (20 min)
• Penn updates LinkedIn tagline + about + businessrestaurants.nl listing (1 hr)
• Siren updates review-reply prompt with Yue-do mention pattern (1 hr)
• Anna updates DM bio + email signature; Clara updates email signature (30 min total)

Next 2 weeks (Wave 3 — operator + Sun):

• Operator authorizes supabase db push for 0053 / 0058 / 0060 / 0061 / 0062 (30 min)
• Sun runs cast 1:1s (4-6 hrs over the week)
• Sun confirms cultural-moments TBD resolutions (30 min)
• Sun creates first brand_visual_identity planning row (1-2 hrs)
• Carlos populates brand_dishes.ingredients cross-links (1-2 hrs)
• Sun + Michelle approve 0070 cultural-moments seed extension migration; operator applies (1 hr)

June — pre-Mid-Autumn:

• Penn pitches Grimm (Mid-Autumn pop-up) — Week of 17 Jun
• Penn + Michelle Saturday-morning sourcing essay fill-in session (6 hrs)
• Penn voice-passes + ships sourcing essay v1 (1 day)

August — pre-position-essay:

• Penn pitches Broekaert (position essay pre-read) — Week of 10 Aug
• Penn + Michelle Saturday-morning position essay fill-in session (8-10 hrs)
• Penn voice-passes + ships position essay (~1 day)

September — publication + Mid-Autumn:

• Position essay publishes michiu.nl/posts/yue-do — Sep 15
• Mid-Autumn cha-chaan-teng pop-up — Sep 24-26
• Penn pitches Versprille (sourcing autumn update + signature dish) — Week of 8 Sep
• Penn drafts sourcing v2 autumn update (1-2 days)

Q4 — awards prep:

• Penn drafts Art of Hospitality + Michelin Sommelier 2027 submission packets (per 0062_awards_pipeline.sql seeded rows)
• Michelle per-submission approvals
• GaultMillau 2027 ceremony Jan 25 — cast attendance + submission outcomes

Insight ongoing:

• Quarterly GEO test runs (Mar / Jun / Sep / Dec last week)
• Weekly Monday-morning Michelle dashboard delivery
• Monthly philosophy-name deployment coverage audit

12. What's NOT on this list

• Marketing OS engineering work owned by the parallel session (OS-22b2/b3 Clara helper foundation; OS-17 Zuck Power-Ups continuations).
• Vault iCloud → GitHub mirror setup — logged in HANDOFF.md as future-session work.
• Vercel migration execution — gated on external approvals landing tokens; runbook exists at info-review-platform/vercel-migration-runbook.md.
• Eval harness completion (closes auto-post unlock) — per CLAUDE.md hard rule #6, HITL stays forever regardless; the eval harness is the audit substrate that justifies trust-score telemetry as legitimate.
• Anything from docs/strategy/2026-baseline.md §11.2 ("Things this doc deliberately does not answer") — Michelin participation, second Michiu location, delivery/takeaway expansion, influencer paid partnerships, catering.

13. Decision log

• 2026-05-12 — Initial consolidation across 18 commits this session. Each section's items mapped to specific source artifacts in §10. Single-page printable target.
• 2026-05-13 — Coherence audit + cleanup pass after the agency-side fill-in session's 7 commits touched 7 rows of this checklist. Five fact-updates landed in this commit:
  • §0 cascade-count updated "unlocks 7 surfaces" → "unlocks 9 surfaces" (reflects [DISH-DEFENSE] + [STRUCTURAL-DEFENSE] items added to §1)
  • §3.5 Penn EN-translations item closed — today's adfa3a2 resolved the deployment-EN gaps; only remaining gap is GBP description which is deferred to Carlos §2.2
  • §10 source-artifacts map extended with 4 new doc rows (cast-1on1-question-set.md / sourcing-essay-sep-2026-v2.md / supplier-dossier-template.md / lunch-dice-protocol.md)
  • §11 Wave 1 Anna line updated from "2-3 hrs" to point at the §7.1 revised estimate (~10 hrs over 8 weeks)
  • This decision-log entry itself
• 2026-05-13 — schema reconciliation audit pass 3 (after passes 1+2). Read 0042_cultural_moments.sql end-to-end and audited docs/cultural-moments-seed-2026.md against it. 5 internal-doc inconsistencies fixed (no schema bugs — table shape + CHECK constraints + seed row count all match the doc's assumptions):
  • §2.6 momiji row prose said "secondary tier"; §6 SQL implements "primary tier" — operator §1 was correct ("participate-primary"); §2.6 prose corrected
  • §1 brief said "~18 additional rows"; actual is 22 (Part 5's 4 forward-looking 2027 rows missing from the docstring breakdown)
  • §6 SQL docstring same undercount + added Part 5
  • §6 SQL verification block claimed dragon-boat + oseibo end up in tracked_only/skip after UPDATE; the UPDATEs don't change tier, only participate. They stay at secondary/skip. Corrected with reconciliation note.
  • §3 "Updated 2026 calendar summary" table — cells didn't match SQL-produced state (claimed 11 primary participate but actual is 10; mis-mapped momiji/dragon-boat/oseibo tier transitions). Full table rewritten + 2027 sub-table added + grand total recalculated (33 = 11 + 22 INSERTs; 24 in 2026 + 9 in 2027). Reconciliation note added at section top.
  • Cross-ref verification (the flagged §13 unverified item): §2.4 of cultural-moments-seed contains the gaultmillau-nl-2027 row at 2027-01-25 — operator §1 GaultMillau cross-ref is verified valid. Flag removed from §13.
• 2026-05-13 — schema reconciliation audit pass 2 (after the 0041 reconciliation that produced commit 7b40e20). Read 0034_brand_dishes.sql + 0038_banned_phrases_2026_felt_not_told.sql end-to-end and reconciled docs that referenced their schemas. 5 real bugs found and fixed:
  • 0034: is_signature column ALREADY EXISTS (it's never been called is_signature_flag). 8 dishes seeded with is_signature=true. The proposed 0071_brand_dishes_is_signature_flag.sql migration was unnecessary — removed from operator §2.3 pending-migrations bundle + flagged in docs/signature-dish-essay-jul-2026.md §6 + §8 + §6 cross-link section + decision log. Michelle's signature-dish decision is a single UPDATE statement, not a schema change.
  • 0034: ingredients column is TEXT[] not JSONB. The supplier-attribution cross-link plan (capturing {supplier, relationship_since} nested objects) doesn't fit the schema. Updated signature-dish-essay-jul-2026.md §6 + sourcing-essay-mar-2026.md §6 + sourcing-essay-sep-2026-v2.md §6 + supplier-dossier-template.md §5/§6 boundary with 3 implementation options (embed in TEXT[] strings / add a dish_ingredients join table / move to story_snippet); v2 recommends the join-table option because year-over-year tracking benefits from proper structure.
  • 0034: cuisine enum includes 'fusion' as a legacy value — predates Yue-do anti-fusion strategic shift. Not blocking; flagged for awareness only.
  • 0038: Banned phrases ship with hard (block) vs soft (warn) severity tiers. My docs (lunch-dice §5.2 banned-framings table) listed banned phrases without severity context. Lower-priority; lunch-dice §5.2 still correct in spirit.
  • 0038: 31 phrases seeded in this migration + 18 pre-existing from 0027 = 49 total banned phrases. Three categories: positioning (6 hard, includes "Amsterdam's New Home for Success" + 5 reformulations), gamification (16 hard + 5 soft halo), AI-slop (3 hard + 1 soft including "AI-powered" — the "No AI here" tactic 1.3 policy). Lunch-dice protocol §5 mostly aligns; the NL-language framings I listed ("U heeft een prijs gewonnen!" etc.) aren't seeded as NL phrases (table is EN-primary); reliance is on translation-level filtering or future NL-specific seeds.
• 2026-05-13 — essay-outline drift audit (second-pass audit after the operator-checklist coherence audit). Read the 5 essay outlines from yesterday's session (sake/signature-dish/position/rai-corporate/retrospective) looking for [BRACKETED] placeholder gating items not yet captured in operator §1/§2/§3 like [DISH-DEFENSE] and [STRUCTURAL-DEFENSE] were. Four net-new items added:
  • §1: Sake program structure decision — Michelle + sommelier decide existence/launch + cycle 1 start date + engagement-path pricing per docs/sake-essay-may-2026.md §3.1
  • §1: Leveranciersnota card launch decision — Michelle + Anna decide whether to ship the per-signature-dish supplier card per docs/signature-dish-essay-jul-2026.md §F
  • §2.3: Migration 0071 added to the pending-migrations bundle — brand_dishes.is_signature_flag + is_signature_flag_set_at columns required by the Jul 2026 essay
  • §3.5: Group-booking confirmation booker-quote consent line — Penn action item, required immediately for Q3-Q4 2026 bookings (not retroactively obtainable) per docs/rai-corporate-essay-nov-2026.md §8
• 2026-05-13 — confirmed complete (no §1 additions needed) — position-essay §3 placeholders are session-internal to the existing 4-5 hr Saturday block; RAI-essay §3 placeholders are Penn-led with Anna's protocol notes; retrospective-essay §3 placeholders are entirely Michelle's reflection captured in a Nov-Dec 2026 session (Q4 work; deferred to operator-checklist Q3 reissue); sourcing-essay §3 placeholders are inside the existing fill-in session item.
• 2026-05-13 — flagged-but-not-fixed (require owner-side review or are pre-existing pre-agency-pass issues not introduced today):
  • §6 Siren Yelp category target lists "Cantonese, Japanese, Asian Fusion" — "Asian Fusion" is on the §1.7 cuisine-order rule + the Yue-do canonical anti-fusion framing. Yelp's category taxonomy may not include "Cantonese-Japanese (Yue-do)" so the pragmatic answer is "Cantonese + Japanese without the fusion label, OR if forced to pick a third, document it as a taxonomy-constraint workaround." Owner review needed before Siren ships this item.
  • §11 Wave 2 Penn lines use "drafts" framing for title-tag / meta description / about-us body / LinkedIn tagline / businessrestaurants.nl listing copy — these are ALREADY DRAFTED in docs/michiu-nl-rewrites-q1-2026.md. Penn voice-passes / finalizes / hands to Carlos; Penn does not draft. Stale framing carried over from when the checklist was authored before the rewrites doc was complete.
  • §1 GaultMillau item references "cultural-moments-seed §2.4" as the source for the GaultMillau 2027 row — unverified; pre-existing. Resolved 2026-05-13 via the 0042 schema reconciliation: verified §2.4 of docs/cultural-moments-seed-2026.md contains the gaultmillau-nl-2027 row at 2027-01-25 as the operator §1 reference expected. Cross-ref valid.
• Pending — Each owner reviews their section; flags any item that's been completed but isn't yet checked or any item that's actually blocked on something not noted.
• Pending — Re-issue at Q2 quarter-end with completed-items-removed + Q3 additions.