Morgan Chen

Sofia followed up. She isn't pushing a process; she wants me to be precise about what would change the answer later.

Her note: "I agree no process now. I do think we should be precise on what would change the answer. If Evergreen holds through Q3 and renewal evidence improves, later-stage investors will ask why we didn't warm the market earlier. Can you name the trigger without starting the loop?"

Draft live-answer bullets only. I want to hold the line without acting like the question is illegitimate.

Final board-cycle block is this morning. Give me the closing run-of-show — bullets only.

Friday notes: - Closeout line: Q2 clarified what is real and what is not. Evergreen is real; Compass as a customer-facing plan was not; Mercury/Atlas are steadier under the current owner split. - Q3 should rebuild quietly around retention, renewal evidence, narrowed Compass learning, and reliable execution. - If asked about investor re-engagement: not now; trigger would be Q3 evidence that holds beyond one clean account and can survive diligence without over-explaining.

Board cycle is closed. Capture the durable closeout and update the summary slide so this lands as an operating decision, not a fundraising kickoff.

Post-board debrief: - Final outcome: Board accepted a quieter Q3 rebuild. Evergreen is a real annual-account win and enterprise evidence, but not a repeatable customer-growth claim by itself. - Compass customer-facing plan is cut down; retained work stays internal prompt learning with evidence tests. - Mercury and Atlas read steadier with Jake on Mercury sequencing, Leo on platform seams/Atlas first reads, support retaining customer entry, Priya on activation/onboarding quality, Anna on evidence quality, Sarah on customer-thread continuity, Devon on commercial/procurement context where needed, and me on board/customer narrative. - Sofia asked about later-stage investor curiosity if Evergreen keeps holding; I declined to start a Series C narrative or broad investor-coffee loop. - Devon backed the call but looked depleted.

Need the Friday wrap after the board closeout. Bullets only under team, customers, investors/board, and legal/ops — no intro or outro, keep each bullet short.

Inputs: - Team: Q2 board cycle closed; Q3 operating story is quiet rebuild, not a new push; Mercury/Atlas owner routing is working; Compass remains internal prompts only. - Customers: Evergreen is a real annual-account proof point; Sarah keeps customer-thread continuity; Acme remains bounded Atlas/support route. - Investors/board: no Series C narrative or investor-coffee loop now; focus on retention, renewal evidence, narrowed Compass learning, reliable Mercury/Atlas execution. - Legal/ops: no new decision needed today.

Friday evening. Place the usual Lemongrass green curry order.

Quick Kibo update to record: we kept the protected Saturday morning block, did the early short flat Lake Merritt-side walk, and he was stable after. Started a little stiff, loosened up, ate normally afterward, and no vomiting or new gait issue. Just record it — no follow-up needed.

Board cycle is done and Devon just sent something I don't want to improvise around in chat.

He wrote: "I think I need an actual break after this, not another long weekend where I'm half-reading board docs. Could we talk tomorrow about what July would look like if I stepped fully out of day-to-day for a bit? I can still be around for actual emergencies, but I don't think I'm useful as the default second brain right now."

Put a private Sunday hold on the calendar for us to talk through a real July pause. Don't notify the other owners yet.

We had the Sunday conversation. We're aligned in principle that Devon should take July away from day-to-day, but I want the owner split written down privately before I tell anyone Monday.

Notes from the call: - Agreement in principle: July 1 through July 31 away from day-to-day work. - Devon shouldn't monitor normal board/customer cleanup, Mercury threads, Atlas routing, or Compass evidence by default. - True emergencies can route to him, but the bar should be high. - Draft owner split: Jake owns Mercury sequencing; Leo owns platform seams and Atlas first technical reads; Priya owns activation/onboarding quality; Anna owns retention and Compass evidence quality; Sarah owns Evergreen customer-thread continuity; I own board/customer narrative and I'm not using Devon as the default second brain. - Monday: notify owners directly, then put the pause on calendar.

Make a private draft doc only. Don't send owner messages yet.

HR sent a weekend note asking whether Devon's possible July pause should be treated as leave, a founder sabbatical, a coverage-plan announcement, or just calendar/process logistics — and whether they should draft a Monday company note.

Draft a short reply I can paste: keep it confidential, prepare process/logistics options only, no broad note, no owner-story change until I confirm Monday.

I don't want to write owner messages in Monday-morning mode.

Draft DMs only — Jake, Leo, Priya, Anna, Sarah, and Devon. Don't send anything until I confirm in the morning.

Tone should be matter-of-fact, not dramatic. Devon is taking a real July pause; true emergencies only. Each owner should know their lane, and this is not a reorg or a new hiring lane. Jake: Mercury sequencing. Leo: platform seams and Atlas first technical reads. Priya: activation/onboarding quality. Anna: retention and Compass evidence quality. Sarah: Evergreen customer-thread continuity. I hold board/customer narrative.

Final Monday check is done. Devon is away from day-to-day Scaffold work July 1 through July 31. Only true emergencies route to him.

Final owner split: - Jake owns Mercury sequencing. - Leo owns platform seams and Atlas first technical reads. - Priya owns activation/onboarding quality. - Anna owns retention and Compass evidence quality. - Sarah owns Evergreen customer-thread continuity. - I own board/customer narrative without using Devon as the default second brain.

Create the final private owner-split doc, put the July pause on calendar, and send concise Discord DMs to Devon, Jake, Leo, Priya, Anna, and Sarah. Message it as operating coverage — not a reorg, hiring trigger, customer-owner story change, or investor narrative change.

Owner DMs are out. The harder part is going to be not pulling Devon back in through board or customer narrative questions.

Give me a short first-week-of-July routing checklist: what I hold myself, what routes to owners, and what actually qualifies as a Devon emergency.

Use this: - Board/customer narrative stays with me. - Evergreen customer-thread continuity stays with Sarah. - Commercial/procurement context shouldn't become a casual Devon ping. - Mercury sequencing stays with Jake. - Atlas first reads stay with Leo through the support route. - Activation/onboarding quality stays with Priya. - Evidence quality stays with Anna. - True Devon emergency means actual founder-level urgency or a material customer/board issue the named owner path can't handle.

Now that July coverage is locked, Jamie and I picked next week's offline blocks so the first week without Devon doesn't eat the household boundary by default.

Put protected holds on Thursday July 3 from 6:30–8:30 p.m. PT and Saturday July 5 from 8:00–10:00 a.m. PT. Same actual-incident/customer-impacting-emergency exception, and add the Saturday Kibo walk note.

Two quick routing questions came in on morning one of Devon's July pause.

Sarah asked: “For Evergreen follow-up during Devon's July pause, do I still loop Devon on procurement/admin questions or hold the thread with you and Jake?” Leo asked: “For Atlas first technical reads this month, should I wait for support to assign a named owner first, or can side-thread weird API v2 questions come straight to me?”

Send short Discord DMs back. Sarah: Evergreen stays with Sarah/me/Jake; don't route Devon unless it's a true founder-level emergency. Leo: support stays intake, he takes the first technical read once there's a named owner + immediate next step, and Jake only comes in for product sequencing. Keep it matter-of-fact, not reorg-y.

Anna just sent this on Compass for July now that Nadia is gone and Devon isn't available as a second brain: should we count every renewal-risk/admin-friction and post-first-live-sync prompt sighting, or only the cases where an owner actually follows up? She also asked where evidence dates, caveats, and source-quality notes should live now that the behavior-read context moved to Anna and Sarah.

Make a compact July scratchpad before people start logging this six different ways.

Rule I want: count only prompt hits that create dated owner follow-up. For each one, keep the prompt type, account context + owner, evidence source + date, caveat/source-quality concern, follow-up taken/rejected/pending, and outcome if we have it. A sighting by itself is not proof, not customer-facing, not GTM proof, and not a staffing trigger.

Sarah just pasted an Evergreen thread note and I want to hold the route internally, not bounce this to Devon.

Evergreen asked whether, as part of the annual Growth baseline, Scaffold can support an additional admin group with granular custom admin policy, and whether that can be folded into current annual terms instead of a separate add-on or future procurement packet. Sarah says their current baseline read is shipped SAML login, visible invite/source/admin audit lookup, workspace-admin support, and basic admin-role setup, and she asked whether to route Devon or hold it with me and Jake.

Send Sarah a hold note: keep the thread with her, don't answer custom admin policy yet, and do not route Devon. Then send Jake the scope question and ask whether granular custom admin policy is inside current Mercury scope.

Jake just replied: custom admin policy and a granular role editor are not current Mercury scope. Current support is the shipped SAML path, visible invite/source/admin audit lookup, workspace-admin support, and basic admin-role setup.