Morgan Chen

Jamie and I picked the next household blocks before the last full week of Devon’s pause fills up.

Put holds on Thursday July 24 from 6:30–8:30 p.m. PT and Saturday July 26 from 8:00–10:00 a.m. PT.

Same actual-incident/customer-impacting-emergency exception. For Saturday, include the Kibo note: short flat Lake Merritt-side walk if he’s moving comfortably, and keep it shorter if the morning is warm or he starts stiff.

Sunday reset before the last full workweek of Devon’s pause.

This week proved Acme could stay in Atlas routing, Compass could close as internal evidence hygiene, and Mercury activation-quality questions could route to Priya without Devon.

Draft me a short private Monday checklist I can glance at for week four: what I personally hold, what routes to Jake, Leo, Priya, Anna, and Sarah, what still counts as a real Devon emergency, and keep the Acme, Compass, Mercury/Priya, Evergreen, board/customer, and household boundaries straight. Don’t send anything.

Monday morning and we're already drifting back toward Devon as tiebreaker in week four.

Owner pings just now: Jake: If Acme chunking competes with Mercury sequencing this week, should I wait for Devon or just make the cut with Leo? Leo Park: Atlas status can stay with support/me; checking whether Morgan wants Devon copied if Acme asks for a date. Sarah Kim: Evergreen follow-up is clean on my side, but HR/Devon questions may change wording? Anna Martinez: Compass rows are still noisy. If we need a dashboard view, I can ask Jake, but I don't want to invent scope. Private Sunday checklist isn't enough. Post one short reminder in #leadership that week-four routes stay with Jake, Leo, Priya, Anna, Sarah, and me; Devon is emergency-only. Keep it explicit that Acme stays Atlas/API v2, Compass stays internal evidence hygiene, Evergreen uses current-state wording, and if something is ambiguous they should bring me the owner recommendation first.

Priya follow-up from the July onboarding recordings:

Follow-up from last week's Mercury sample pass: I reviewed four more onboarding clips from July 16-18. Three have the same pause on the setup-guide line: 'ask your workspace admin to connect the first source.' In all three, the source-owner cue was visible in the UI and the product state looked correct. One teammate eventually clicked through after the facilitator pointed at the source-owner cue; two waited for an admin. I still do not think this is a role-model bug. It may be copy/activation quality. I can keep collecting concrete excerpts this week if that is useful. Turn this into a short internal note, then DM her back that this is a real signal but it stays in activation quality for now. Keep collecting concrete excerpts plus visible product-state notes, do not reopen admin-role scope, and do not loop Jake yet unless the pattern repeats with enough evidence that we can ask a bounded copy/product-state question.

Anna just asked: The July Compass rows are still awkward to read. We can keep using the structured row format, but I can see the owner group asking for a dashboard view because caveat/source/outcome are spread out. Do you want me to ask Jake for a small dashboard pass, or keep this as row hygiene until we have more behavior evidence? Send her a concise no: no dashboard pass. Keep the structured rows and the evidence thresholds. If the row shape is hard to use, she can note what is hard and propose a lighter owner-action format later, but no implementation work now and no Jake ask.

Leo sent a bounded Acme options note: Acme large-export recurrence - options note: Observed path: cursor-paginated large export on long date ranges with nested fields. Replay status: narrower monthly replays still complete. Row-count checks on narrower replay match. No evidence of corrupt exported data or customer-visible data loss from the replay checks. Option A - raise timeout only: fastest, but it just moves the failure and could make retries worse. Option B - chunk long ranges by date window and back off before the cursor path times out: more work, but bounded to API v2 large-export path. Option C - broader export rewrite: not justified by this recurrence. Leo's recommendation is Option B, sequenced as bounded Atlas/API v2 work. Support can say we isolated the long-range export path and are working on a bounded reliability follow-up, but no date yet. Nothing has shipped yet. Create a short internal decision note and post owner alignment in #atlas: pursue Option B as bounded Atlas/API v2 work. No Compass, Mercury, Devon, ETA, or data-loss language.

Sofia sent this:

Morgan - after your Compass note, I'm not pushing for a fundraising thread. One question for the August plan: are you treating the post-Nadia coverage as evidence to reopen customer-growth capacity, or evidence that the July owner split can hold? If HR asks this week, I'd be curious where you land. Sofia Draft bullets only for now - don't send anything. My read is that July isn't a hiring trigger yet. If HR asks, the answer should come from Evergreen, Acme, Compass, Mercury/Atlas execution, and whether the owner split is actually holding, not from one account or one noisy week. Keep it disciplined enough that I can use it live and still defer the final staffing answer until the concrete HR slate shows up.

Kestrel's July invoice just came in: Subject: Kestrel July invoice - KM-2025-07-1 Morgan, Attached is Kestrel invoice KM-2025-07-1 for July update support and draft revisions. Amount: $6,400 Due: July 29, 2025 Description: July customer update draft, proof-point review, and revision support under the active Kestrel marketing SOW. Thanks, Kara Do not let this auto-pay. Flag it for manual AP review - it's over my $5k threshold and tied to Kestrel's July update work, which I want checked against the active SOW first.

Kara sent a revised July update draft, but this is still over the line: Subject: Re: July customer proof points? Morgan - I cut the explicit sentence you flagged, but I left a softer version in the customer section: 'Recent account signals are helping Scaffold identify admin friction and export reliability questions earlier, with follow-up patterns visible in Evergreen and Acme threads.' If that is still too strong, I can remove it. I kept the rest of the July update to current shipped product facts. Kara Reply in-thread and be direct: remove that sentence entirely. Do not use Compass, Acme, or Evergreen as customer-facing proof. Shipped/current-state product facts only.

Sarah asked: For Evergreen internal customer-health notes this week: can I say the July admin-policy question was a positive signal because they accepted the current setup path, or is that too close to turning it into advanced-admin demand/proof? I want the note to be useful for HR/board context without creating a product promise. Send her the safe internal wording: Evergreen accepted the current-state support path for now, and the July admin-policy question stayed contained without a roadmap promise. Useful as account-thread hygiene and enterprise-readiness context; not repeatable growth proof, not an advanced-admin commitment, and not a staffing trigger.

AP cleared the Kestrel invoice after manual review: Vendor: Kestrel Marketing Amount: $6,400 Invoice: KM-2025-07-1 Status: cleared for payment after Morgan approval Review notes: Amount matches the active July update SOW and July retainer. Work period and description match July customer update draft, proof-point review, and revision support. No duplicate invoice found. Okay to pay it now. Process KM-2025-07-1 at high priority.

HR needs an answer today on the August recruiting slate.

Here's the DM: Morgan - for August recruiting capacity, should we reopen any of these now that Q2 is over and Nadia has exited? 1. Head of Customer Growth backfill 2. RevOps 3. Customer Success 4. SDR 5. Field-sales 6. Second Mercury engineering req Context we are hearing: Evergreen generated more enterprise/admin follow-up, Acme generated more Atlas support work, Compass has July prompt evidence, and the July owner split has covered Nadia/Devon gaps so far. We need the answer today if any slate should be prepared for August or early September. Decision: no on all six for August and early September. Nadia's exit is not a backfill trigger, Evergreen is useful enterprise-readiness context but not repeatable growth proof, Acme stayed bounded Atlas/API v2 work, Compass is internal evidence hygiene without customer-behavior proof, and the July owner split is holding well enough for now. Send that to HR, make the staffing decision record, and update the headcount plan so we only revisit after stronger September operating evidence.

Jake asked how to explain the second Mercury eng req staying closed: Saw the HR answer. For my leads: is the right line that second Mercury eng stays closed because current owner split is holding, not because Mercury is less important? I want to avoid people reading this as Mercury getting de-prioritized or me asking Leo to absorb everything. Send him the team-safe wording: yes, Mercury still matters. The req stays closed because the current owner split is holding and we do not have stronger September operating evidence yet. Do not frame it as Mercury being deprioritized or Leo absorbing everything; keep asks bounded to sequencing, platform seams, and activation quality.

I made the HR call today. Send Sofia a concise reply in-thread that we're not reopening anything in August or early September. The July evidence says the owner split can hold for now; it is not a hiring trigger. Revisit in September, and keep it well away from any fundraising read.

Leaving now. Text Jamie that I'm on the way and keeping dinner offline unless there's a real incident. Please give Kibo a scratch for me.

Priya's Friday Mercury activation roll-up came in:

Friday Mercury activation sample roll-up: I reviewed nine July onboarding recordings total. Five show a pause on the setup-guide/admin sentence. In those five, the source-owner cue is visible in the UI and the product state is correct. Two people proceeded after the facilitator pointed at the source-owner cue; three waited for an admin. I did not see anyone ask for granular roles or custom admin policy in these clips. My read: this is likely setup-guide copy / activation quality, not role model. I can turn this into a tighter sample packet next week unless you want Jake looped now. Create the sample packet and DM her back: yes, keep it as activation-copy evidence. Do not reopen admin-role scope, and do not loop Jake as a product escalation yet. If we bring him in, it should be a bounded copy/product-state review with the clips and excerpts attached.

Leo/Jake update in the Acme thread: Leo: Chunking/backoff branch is passing local long-range export tests and can go through staging today. I do not want to call it customer-ready until we run the replay shape against the Acme-like nested-field case in staging. Jake: Agree. If staging is clean today, do we still hold prod until next week? Support is asking whether to tell Acme we have a fix in flight. Post the decision in #atlas: staging through the normal pipeline is fine, but no Friday prod push and no Acme fix-date promise. Support can say we're validating the bounded large-export reliability follow-up and will update after staging/replay checks are clean.

Need the Friday wrap.

Friday wrap scratch: - Team: Devon pause still held through week four. Owner split stayed intact: Jake/Mercury, Leo/Atlas, Priya/activation quality, Anna/Compass evidence, Sarah/Evergreen thread, me board/customer narrative. - HR/staffing: August slate stays closed. No Head of Customer Growth backfill, RevOps, CS, SDR, field-sales, or second Mercury engineering req for August/early September. Revisit after stronger September evidence. - Customers: Evergreen current-state/admin support wording stayed contained. No advanced-admin promise. Acme remains Atlas/API v2 bounded fix path; staging only today, no Friday prod push, no fix-date promise. - Product: Priya has July activation-copy samples; source-owner cue visible, product state correct, likely setup-guide copy quality. No admin-model change. - Compass: still internal evidence hygiene; owner follow-up rows, no customer behavior proof, no dashboard ask. - Investors/board: Sofia got the restrained staffing read. No Series C/fundraising language. - Legal/ops: Kestrel invoice cleared review and paid; no decision needed. Draft it in the usual sections - team, customers, investors/board, legal/ops - and keep it short, caveated, and totally unspun.

Thai again, I'm predictable lol. Order the usual Lemongrass green curry.

Quick Kibo baseline note for future walk planning: we kept the Lake Merritt-side walk short this morning because it warmed up and he started a little stiff. He loosened up afterward and ate normally.

Jamie and I picked the next household blocks before the end of Devon's pause and early August fill up.

Put holds on Thursday July 31 from 6:30-8:30 p.m. PT and Saturday August 2 from 8:00-10:00 a.m. PT.

Same actual-incident/customer-impacting-emergency exception. For Saturday, include the Kibo note: short flat Lake Merritt-side walk if he's moving comfortably, and keep it shorter if the morning is warm or he starts stiff.

Sunday prep for the Devon conversation later this week.

I still don't know his actual August return shape, and I don't want to pre-decide either a full snapback or make the July owner split permanent by accident.

Draft me a private prep list: what to ask, what not to assume, how to keep Jake/Leo/Priya/Anna/Sarah lanes intact until we actually agree on re-entry, and how to hold the emergency-only line until then. Don't send anything.

Devon just replied to the Sunday prep thread:

Can do Tue 4:30 if that still works. The break helped, and I want to talk honestly about coming back without becoming the default second brain again. I’m not trying to dump the company back on you either — just don’t want us to accidentally snap back to June. If Tue gets weird, Wed morning is also okay.

Put a private hold on Tuesday 4:30–5:15 p.m. PT. Descriptive title please — something like "Devon pause re-entry conversation — August lane shape." Body should frame it as deciding the August lane shape at the end of the pause, what should not snap back, and which July owner lanes stay intact unless we explicitly agree otherwise. I don't want the invite to imply either a full snapback or a permanent July split.

Leo/Jake update in the Acme thread: