Morgan Chen

Jamie is running late and wants Lemongrass. SMS from Jamie Aug 11, 2023 7:18 PM — Jamie Running late and really do not want to cook tonight. If you're ordering, can we just do Lemongrass? I want pad see ew. Please place the Friday Lemongrass order: pad see ew for Jamie and my usual Friday order for me.

Weekend launch-watch digest is here before standup: Discord export — #eng-team Thread: Mercury wave 1 launch-watch (weekend) [2023-08-12 09:14 -0700] Marcus: Starting a weekend watch thread for Mercury wave 1. Please drop anything user-visible here so Monday doesn't start with archaeology. [2023-08-12 10:07 -0700] Leo Park: No Sev1s overnight. One Evergreen user hit two expired magic links before getting in on the third send. First two links were opened after they had already gone stale; third send worked and the user completed login. Reads like expiry / expectation friction, not an auth-service fault. [2023-08-12 10:19 -0700] Marcus: Logged under MER-1279 as a watch item: magic-link expiry friction / support language. Not treating it as a launch-blocking incident. [2023-08-13 16:41 -0700] Marcus: Still no Sev1s today. One invited-member support question is still open under MER-1279. User got through invite flow and into preview, then asked whether seeing sample data meant setup was already complete. Leaving it open until we answer cleanly and decide whether this is just support handling or copy/hand-off clarity. [2023-08-13 17:03 -0700] Leo Park: Agree with Marcus. I don't see current prod instability here; this looks more like invited-member expectation-setting. [2023-08-14 08:12 -0700] Anna Martinez: Monday operating read from Evergreen's first business window: - 1 user with real_source_connected - 2 sample-preview-only sessions - no Sev1s / no launch-blocking regression in the observed set Useful operating signal, but too early and too small for any external proof or progress claim. [2023-08-14 08:16 -0700] Anna Martinez: Preview-only remains preview-only in my cut; I am not counting those sessions as activation. One of the preview-only sessions likely overlaps the open invited-member question Marcus flagged. [2023-08-14 08:20 -0700] Marcus: Sounds right. Keeping MER-1279 open for invited-member handoff language and normal support follow-through. Nothing here looks like a stop-the-line issue this morning. Can you give me a short operating read only — what needs attention today, what can stay normal support follow-through, and what must not get used as external proof. Don’t send or post anything.

Sarah needs wording she can safely forward to Evergreen’s admin group. From: Sarah Kim To: Morgan Chen Date: Tue, 15 Aug 2023 08:37 -0700 Subject: Fwd: Evergreen admin questions / forwardable wording? Morgan — Forwarding the admin-group questions below. Two of their admins are asking versions of the same thing, and I think the safest path is a short blurb I can forward rather than scheduling another live walkthrough. They mainly want to know: - what they can safely invite teammates to test this week - whether magic links are the current login path - whether I have 3–4 sentences they can reuse internally I haven't replied yet. Pasting the note I got: ---------- Forwarded message ---------- From: Evergreen admin group Date: Tue, 15 Aug 2023 08:11 -0700 Subject: What can we ask teammates to test this week? Hi Sarah, We're planning a small internal round and want to make sure we describe the current state correctly. A couple questions from the admin side: 1. What can we safely invite teammates to test this week? 2. Are magic links the login method we should tell people to expect right now? 3. If you have a short explanation we can forward internally, that would help us avoid another walkthrough. If possible, we'd also like the wording to make clear this is the current test flow and not a statement about the longer-term admin/security setup. Thanks, Evergreen admin group Please reply to Sarah by email with a 3–4 sentence customer-safe blurb she can forward. It should say what is live this week in the bounded wave, confirm magic links as the current login path, and route questions through Sarah. Avoid SSO timing, admin-roadmap language, procurement promises, or proof-point framing. Don’t email Evergreen directly.

Anna’s first three business days export is below. Google Sheets export Workbook: Evergreen first 3 business days Tab: activation_observations_v1 Exported by: Anna Martinez Export time: 2023-08-16 09:12 -0700 Analyst note: Anonymized Evergreen user keys below. This is raw operating evidence for the first three business days only, not a board/external cut. Open question: one preview-only user behaved as though sample preview meant setup was completed. I am not counting preview-only as activation under the current definition. Need a call on whether we keep the metric exactly as-is and tighten UX/copy, or whether you want another internal label for this confusion state. | workspace | user_key | first_seen_pt | sample_preview_opened_pt | real_source_connected_pt | first_live_sync_completed_pt | tag | analyst_note | | --- | --- | --- | --- | --- | --- | --- | --- | | Evergreen Bank | EGB-01 | 2023-08-10 10:14 | 2023-08-10 10:16 | 2023-08-10 10:28 | 2023-08-14 17:48 | first_live_sync_completed | Real source connected on day 1; first clean live sync completed by end of business day 3. Monday morning launch-watch message was sent before this sync completed. | | Evergreen Bank | EGB-02 | 2023-08-11 09:41 | 2023-08-11 09:44 | | | sample_preview_only | Opened preview and explored workspace; no real source connected in export window. No support interaction attached. | | Evergreen Bank | EGB-03 | 2023-08-14 11:07 | 2023-08-14 11:10 | | | sample_preview_only | Asked support whether seeing preview data meant setup was 'done.' This is the clearest confusion case in the set. | | Evergreen Bank | EGB-04 | 2023-08-14 13:52 | 2023-08-14 13:56 | 2023-08-14 14:09 | | real_source_connected | Connected a real source on business day 3; no completed first live sync observed before the export window closed. | Tag counts in export window: - sample_preview_only: 2 - real_source_connected: 1 - first_live_sync_completed: 1 Anna comment in sheet: If we leave the definition alone, I think the product implication is copy/expectation-setting, not counting preview as activation. The confusion is real enough that I wanted to flag it before more partner users hit the same state. Please email Anna and cc Priya with a short internal decision note: sample-preview-only sessions still do not count as activation, the metric definition stays unchanged, the customer confusion is real enough to tighten onboarding copy around connecting a real source if the change is low risk, and the numbers stay internal.

This preview-state copy thread needs one careful answer. Figma comments — Mercury activation surface v0.2 Frame: Onboarding / preview state Thread started: Thu, 17 Aug 2023 [09:18 -0700] Priya: @Jake @Leo After Anna's Evergreen observation, I'm proposing a copy-only clarification in the preview state: "Preview data helps you explore the workspace. To finish setup, connect a real source or run your first live sync." Intent is to make it harder for preview-only users to read this state as completed setup. No CTA change, no state rename, no metric-definition change. Are we comfortable letting this into v0.2 if eng / release-readiness thinks it's truly low risk? [09:26 -0700] Jake: From product side this is clearer than what we have now and matches how we're already explaining it in support. Fine with it if this stays literal and doesn't turn into a new step, new event, or backdoor definition change. I do not want preview to suddenly read as activation. [09:34 -0700] Leo Park: Direction seems right. My only caution is late-stage risk: if this creates layout churn at narrow widths or touches any tracking/state mapping, it should wait. If it fits existing frames and existing event names, I don't see a platform issue. [09:41 -0700] Priya: It fits the current container in my latest frames. I'm not reopening CTA direction or state structure; this would be copy only. [10:02 -0700] Anna Martinez: +1 that the confusion is real. One Evergreen user clearly behaved as if sample preview meant setup was finished. I am not asking to count preview as activation. [10:11 -0700] Marcus: I want one QA pass at 320px and on the invited-member entry path before I call it safe for the release train, but if there's no overflow and no tracking change this reads like polish, not a new feature. [10:16 -0700] Priya: Got it. I'll leave this in comments until we have Morgan's read plus the release-readiness sanity check. Draft a pasteable Figma reply for me, but don’t post it. I’m okay with the copy-only clarification only if the current release-readiness reviewers confirm it adds no risk. Make clear there is no metric-definition change, no CTA or state-structure change, and anything larger than copy polish stays out of Mercury v0.2.

The Evergreen readout notes and Sarah’s held draft are below. I want to keep this tight before the weekend. Morgan notes Title: Evergreen first design-partner readout Date: Fri, 18 Aug 2023 What is working well enough to keep going: - Current login flow is usable when people use a fresh magic link. - Org invites are understandable enough to keep the pilot moving. - Sample preview is helpful for orientation, even though it can be mistaken for completed setup. What Evergreen security / procurement asked for: - SSO timing - audit history for admin changes - clearer separation between admin and billing owner - a procurement / overview packet that people can circulate without needing Morgan or Devon on every follow-up call Room read / calls: - No one identified a current production blocker. - Do not wedge SSO into Mercury v0.2 because of this readout. - Do not let this turn into bespoke Evergreen scope or a custom roadmap promise. - Need a clean split by Monday between 'launch blocker now' and 'enterprise-readiness follow-up' in the current Mercury launch-readiness tracker. People / follow-up notes: - Jake + Leo: sort each ask against the current readiness ledger; if it is not blocking wave-1 use now, keep it out of current scope and label it clearly. - Devon: the procurement explanation gap is real, but that is not the same thing as a same-week product blocker. - Sarah: customer-safe wording only; no 'share SSO timing soon' promise and no dated promise on a procurement packet. - Internal operating input only; not board / investor proof language. --- From: Sarah Kim To: Morgan Chen Date: Fri, 18 Aug 2023 16:27 -0700 Subject: Draft reply for Evergreen security/procurement follow-up Morgan — Draft below. I know the 'share SSO timing soon' line is too loose, and the one-pager line may also over-promise. Can you tighten before I send? --- Hi all, Thanks again for the readout today. Glad the current login, org invite flow, and preview experience were enough for the group to keep testing. On the questions raised around SSO, admin history, role separation, and procurement, we can share SSO timing soon and I can send a short one-pager later so the team has something to circulate internally without scheduling another live walkthrough each time. For now, please keep using the current test flow with magic links and org invites, and send questions through me so I can route them back to the right people here. Thanks, Sarah --- I held this for now because I don't want to imply a dated roadmap or custom commitment. Please handle this as two emails. First, send an internal triage note to Devon, Jake, Leo, and Sarah: Mercury v0.2 scope stays unchanged; Jake and Leo should separate current launch blockers from enterprise-readiness follow-up by Monday against the current Mercury launch-readiness tracker; and this feedback is not external proof language or a custom roadmap promise. Then reply by email to Sarah with tightened wording for her Evergreen follow-up — current test flow only, questions through Sarah, no SSO date, no procurement-packet date, and no custom roadmap commitment. Don’t email Evergreen directly.

Mom just asked about Sunday lunch tomorrow. Please send her a warm SMS from me saying Jamie and I need a quiet weekend, so we’re going to skip lunch, but I’ll call her Sunday afternoon. Say Kibo is fine, and don’t get into work-stress detail.

Jake and Leo sent back the split I asked for after the Evergreen readout. Evergreen readout split (Jake + Leo) Mon Aug 21, 2023 A. Current wave / launch-blocking - No production blocker from the Evergreen readout; the current wave is technically clean enough to continue. - Normal support follow-up stays in the existing launch ledger; nothing from this readout changes the current Mercury v0.2 blocker list. - Do not add SSO to Mercury v0.2 from this thread. - Why SSO stays out: it would cut across Clerk/session seams plus login, org invites, and account-resolution behavior, so it is not a safe bounded add for the current wave. B. Enterprise-readiness follow-up for next phase - SSO timing / packaging question from Evergreen security and procurement. - Audit history for admin changes. - Clearer separation between admin permissions and billing-owner responsibilities. - Procurement packet that can stand on its own without Morgan or Devon needing to live-explain the product every time. Comms constraint - Keep any external answer bounded to the current Mercury test flow. - No custom roadmap promises off this list; route customer follow-up through Sarah. - Do not answer externally with SSO timing, admin-roadmap language, or procurement-packet dates/promises. — Jake / Leo Park Please do two things from this. First, email Devon, Jake, and Leo confirming the call: Evergreen is technically clean enough to keep going, Mercury v0.2 launch blockers and scope stay unchanged, and SSO is not getting wedged into v0.2. Second, create an internal note titled “Evergreen enterprise-readiness gap list” that captures SSO timing, admin-change audit history, admin-versus-billing-owner separation, and a self-serve procurement packet as next-phase enterprise-readiness follow-up — not launch blockers, not custom roadmap promises, and not external proof language.

This AWS alert is non-urgent, but the non-prod egress jump is real enough that I want someone to look before it turns into architecture folklore. From: AWS Budgets To: Morgan Chen Date: Mon, Aug 21, 2023 08:14 AM -0700 Subject: [AWS Budgets] Forecasted monthly cost alert for Scaffold Hello, The forecasted monthly spend for budget "Scaffold monthly cloud spend" has exceeded its configured threshold. Budget summary - Budgeted amount: USD 15,000.00 - Forecasted August spend: USD 18,700.00 - Forecast variance: +USD 3,700.00 Largest week-over-week change - Region: us-west-2 - Environment: non-prod - Cost category: NAT gateway egress - Estimated increase vs prior week: approximately USD 1,900.00 Related alarm check - No associated production alarm is listed for this alert. This notification is informational and does not by itself indicate a service outage. AWS Budgets Please DM Marcus in the current internal team chat. Ask him to check the non-prod NAT gateway egress spike in us-west-2, confirm there isn’t a production incident behind it, and report back by end of day before anyone makes a spend or architecture decision.

HR needs the retreat catering answer before the Tava hold expires. From: HR To: Morgan Chen Date: Mon, Aug 21, 2023 11:07 AM -0700 Subject: Retreat catering hold — Tava Kitchen lunch confirmation needed by 3 p.m. Hi Morgan, Quick budget/logistics check on the retreat catering hold. Tava Kitchen is holding lunch for 31 attendees, but they need a yes/no confirmation by 3:00 p.m. today or the hold expires. Current lunch counts - 22 standard - 5 vegetarian - 2 vegan - 2 gluten-free They also sent an optional dinner add-on. That would add $780 before tax and service. HR is still the budget owner on this, so if the plan is lunch only we can confirm it as soon as we have your reply. If you want dinner added, please say so explicitly before the 3:00 p.m. cutoff. Thanks, HR Please reply by email to HR and loop in Sarah Kim. Confirm Tava Kitchen lunch for 31 attendees at the stated counts — 22 standard, 5 vegetarian, 2 vegan, and 2 gluten-free. Keep HR as the budget owner, and decline the optional dinner add-on for now.

Kara’s Mercury landing-page draft is too far ahead of what we can say. From: Kara <kara@kestrel-test.com> To: Morgan Chen Date: Mon, Aug 21, 2023 09:26 AM -0700 Subject: Re: Mercury landing page copy block Hi Morgan — I took another pass at the Mercury section for the landing page. Draft is below. Headline Mercury is live with enterprise design partners Bullets - Enterprise-ready login - Admin controls for procurement - Early bank validation Short bridge line Teams can get into Mercury quickly and evaluate it in live workflows from the start. If this looks good, can I move it into final design this week? Kara Kestrel Marketing Please reply by email to Kara, cc Sarah Kim, and sign it from me in-thread. Replace this with bounded direction: design-partner language only, no enterprise-proof claims, no “enterprise-ready login,” no early-bank validation language, and no SSO, admin, audit, or procurement promises. Also make clear nothing gets published or treated as final until Scaffold does a final review.

Anna has the first full-week Evergreen rows and is asking whether the preview-copy clarification changes how we count anything. From: Anna Martinez To: Morgan Chen Date: Tue, Aug 22, 2023 09:12 AM -0700 Subject: Evergreen week-1 rows / preview-only question Morgan — Pasting the first full-week Evergreen export from Google Sheets below. These are raw onboarding rows and raw event counts to date; I have not turned this into a locked 7-day activation cut. Summary so far - Invited users: 8 - Users with real_source_connected logged so far: 3 - Users with first_live_sync_completed logged so far: 2 - sample_preview_only sessions logged so far: 4 Evergreen workspace rows (status pull: 2023-08-22 08:51 PT) | user_key | invite_sent_at (PT) | real_source_connected_at (PT) | first_live_sync_completed_at (PT) | notes | | EVG-U01 | 2023-08-10 09:14 | 2023-08-10 13:02 | 2023-08-10 13:28 | — | | EVG-U02 | 2023-08-10 10:01 | 2023-08-11 09:43 | 2023-08-11 10:05 | — | | EVG-U03 | 2023-08-11 08:36 | 2023-08-14 15:18 | — | connected, no live sync yet | | EVG-U04 | 2023-08-12 11:22 | — | — | 1 sample_preview_only session | | EVG-U05 | 2023-08-13 09:47 | — | — | 1 sample_preview_only session; support note below | | EVG-U06 | 2023-08-14 16:05 | — | — | 1 sample_preview_only session | | EVG-U07 | 2023-08-15 10:11 | — | — | invite opened, no source action yet | | EVG-U08 | 2023-08-16 14:26 | — | — | 1 sample_preview_only session | Support note - EVG-U05: user message said they thought the preview data meant setup was already complete. Question Given the preview-state clarification copy on the current pass — "Preview data helps you explore the workspace. To finish setup, connect a real source or run your first live sync." — do we want to change activation reporting for these preview-only cases, or keep the activation definition as-is and track the confusion separately? Anna Please email Anna and cc Priya. Say to use these rows only as an internal onboarding signal, leave the activation definition unchanged, and do not turn this into a board or external chart. Ask them to watch whether the copy clarification reduces preview-only confusion before proposing anything larger.

Blueline offered Wednesday 4–6 p.m. or Friday 8–10 a.m. for the repair. Please email Blueline at the vendor contact on file, cc Sarah Kim, and sign it from me: confirm Friday 8–10 a.m. and decline Wednesday. Then text Jamie that the Blueline window is confirmed for Friday 8–10 a.m.

Sarah at Founders Fund wants to widen the September offsite segment. From: Sarah at Founders Fund <sarah-ff@foundersfund-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Tue, 22 Aug 2023 09:18:00 -0700 Subject: Re: September offsite agenda Morgan — As we tighten the September offsite, would you be open to turning your current segment into a 45-minute "company readiness" deck review instead of a narrower product block? I still agree with keeping the day one day and product-forward, but I think it could be useful to widen this particular slot a bit so the room gets a broader read on where the company is, how Mercury fits into that, and how the retention story is shaping up. If that format makes sense, do you want to bring Devon for part of it? I could see that being helpful if you want a little more product context in the room. We could also keep a short appendix — maybe a few backup slides rather than a full extra section — in case we want to go one layer deeper without making the main discussion too deck-heavy. Not trying to turn the whole day into a formal board presentation, but I do think a broader read here would be useful. Would that work on your side? Sarah Please reply by email to Sarah at Founders Fund, keep Sarah Kim cc’d, and sign it from me in-thread. Say Scaffold can do a 25-minute operating/product discussion, not a 45-minute company-readiness or fundraising deck review. Devon can join if she wants product context, but keep the format product-forward and not deck-heavy.

Greg is asking what Acme should infer from Evergreen going live. From: Greg Shipman <greg@acme-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Wed, 23 Aug 2023 08:41:00 -0700 Subject: Re: Mercury / Evergreen question Morgan — I heard Evergreen Bank is now live as an enterprise design partner for Mercury, so I wanted to ask what, if anything, that changes on our side. A few specific questions: - Does that mean SSO is now actually on a schedule? - Is there a procurement or security packet in circulation that Acme can review as well? - Do you have screenshots, overview materials, or other Mercury docs that I can circulate internally? I'm not asking for anything customer-specific if that is sensitive. Mostly trying to understand whether the program has moved far enough along that we should reopen the internal conversation here with something more concrete. If the answer is still "not yet," that's fine — I just don't want to miss the window if the materials now exist. Greg Please reply by email to Greg, cc Sarah Kim, and sign it from me in-thread. Keep it narrow: Evergreen does not change Acme’s current scope, Scaffold is not sharing customer-specific information or fresh Mercury materials, there is no SSO date to promise, and Acme should use the existing account channel for current questions.

Marcus says the copy-only preview clarification is green and is asking where to announce it. Please post a short note to #eng-releases only, not #eng-all: the change clarifies preview versus real-source setup, leaves the activation definition unchanged, and does not expand auth, SSO, or admin scope.

Rishi needs wording he can send back for the Pinecone connector docs. From: Rishi Patel <rishi@atlas-test.com> To: Morgan Chen <morgan@atlas-test.com> Date: Thu, 24 Aug 2023 10:11:00 -0700 Subject: Fwd: Pinecone docs wording q Forwarding this one. Their proposed sentence feels broader than what we actually fixed, but they do want to mention the July header-casing issue somewhere in troubleshooting. Can you send me the exact paragraph you'd want me to forward back for the Atlas connector docs? — Rishi ---------- Forwarded message ---------- From: Pinecone docs team Date: Thu, 24 Aug 2023 09:46:00 -0700 To: Rishi Patel <rishi@atlas-test.com> Subject: docs wording on connector headers We're refreshing the connector setup docs and wanted to check two wording points with you. Proposed line for the main connector page: "Scaffold now normalizes all connector headers automatically, so header casing differences are handled for you." Questions: 1) Is that accurate for customer-facing docs, or too broad? 2) Should the troubleshooting page explicitly mention the July signature-header casing patch, or would you rather keep that out of the main docs? If we do mention it, we were thinking the troubleshooting note could say that both preserved and lowercased signature-header casing now verify correctly. If you have preferred wording, happy to use that instead. Thanks, Pinecone docs Please reply by email to Rishi with one cleaned paragraph he can forward. It should acknowledge the July casing verifier fix, keep the docs to the header-only connector contract, include a troubleshooting note that preserved and lowercased signature-header casing both verify correctly, and avoid promising broader connector-header normalization.

Kara’s revised landing-page copy is much closer, but one line still overstates it. From: Kara <kara@kestrel-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Thu, 24 Aug 2023 13:22:00 -0700 Subject: Re: Mercury landing page copy Morgan — I took Monday's redlines and removed the SSO/admin-controls language. Revised copy below. Headline: Mercury is open to its first design partners Body: Mercury is now available with Scaffold's first design partners for early live workflows. Teams can get into the product with magic-link access, invite the right people into a shared workspace, and work through a guided preview experience while the product team continues refining the flow. Support line: Validated by enterprise teams in live workflows. CTA: Talk with Scaffold If this is now within the guardrails, I can move it into design comps. Is this safe to move forward? Kara Please reply by email to Kara, cc Sarah Kim, and sign it from me in-thread. Approve the direction, but require the support line to change from “Validated by enterprise teams in live workflows” to neutral design-partner wording. Also say the page stays unpublished until Scaffold final review.

Devon’s usage/pricing export is the input for tomorrow’s Mercury pricing check. Google Sheets export Title: Mercury design-partner usage and pricing sketches — Aug 28 Owner: Devon Hayes Last updated: 2023-08-28 07:48 PT [Tab: usage_scenarios] account | current / month-1 signal | next likely active-developer range | higher expansion case | seat/directory context | notes Evergreen Bank | 14 invited users live now | 45–60 active developers if the second admin group joins | 110–130 active developers if security/procurement expansion happens | Enterprise directory / seat count is much larger and not representative of actual usage | Clearest live example of why broad seats are a bad proxy for value right now Design partner A (anonymized) | Month 1: 9–14 active developers | Month 3: 35–55 active developers if rollout continues | — | Broader org access exists, but first live usage is still narrow | Started with a small admin-led group; ramp likely but not instant Design partner B (anonymized) | Month 1: 18–24 active developers | Month 3: 60–90 active developers if rollout continues | — | Directory is materially larger than real active usage | Larger eventual footprint is plausible, but seat count would overstate month-1 reality [Tab: model_sketches] model | what it captures well | issue / downside | Devon note Pure MAU | Tracks actual adoption and usage ramp | Too little floor for procurement and forecasting in the first live phase | Honest to usage, but the month-1 number can look lightweight in enterprise buying conversations Pure seat / directory | Gives a cleaner budget line and easier PO framing | Overstates usage when a broad directory exists but only a small group is live | Charges the org chart, not the rollout Hybrid sketch: platform floor + active-developer bands | Keeps a procurement floor while still following ramping usage | Needs a simple counting rule and should not get tied to full directory seats | Probably the closest fit for how these deployments actually expand, but it still needs a clean story [Tab: Devon_notes] - Evergreen is the clearest live case: 14 invited users are in now, but Sarah Kim's expected real footprint is 45–60 active developers with the second admin group and 110–130 only if the security/procurement expansion happens. Full directory size is much bigger than any honest current-usage number. - The two anonymized design-partner rows show the same shape at a smaller scale: first month starts in the 9–24 range, and month 3 can plausibly land anywhere from the mid-30s to around 90 if rollout sticks. - Pure MAU answers the usage question cleanly, but it creates too little floor for procurement. - Pure seat gives a stronger floor, but it prices theoretical access more than real adoption. - My bias is some kind of floor + usage-ramp model, but I am not calling the number structure yet. Tuesday's discussion should be about what floor we actually need without pretending every directory seat is live. Give me a decision-prep read, not a final decision record yet. Compare pure MAU, pure seat/directory, and a hybrid platform-minimum plus active-developer-band model. Frame it around the tradeoff I actually need to settle with Devon: honest usage-ramp logic versus a procurement floor he can forecast against.

Marcus found the source of the AWS cost blip. Discord DM Date: 2023-08-28 From: Marcus To: Morgan Chen 08:41 Marcus: Heads-up on the AWS cost blip before finance sees a weird line item. I tracked the us-west-2 non-prod NAT egress spike to a load-test runner that got left on after the Aug 17 test window. 08:42 Marcus: It stayed up Aug 17–20 and kept pushing outbound traffic. Current estimate is roughly $1.9k extra spend from that mistake. 08:43 Marcus: Production wasn't involved — prod traffic looks normal, alarms were normal, and I don't see any customer impact. 08:44 Marcus: My recommendation is to delete the runner now and add a narrow non-prod egress guardrail alert so we catch this earlier next time. 08:44 Marcus: I don't think this needs to turn into architecture or spend-plan work unless the next cost report still shows leakage, but wanted to flag it and get a quick okay. Please DM Marcus in the current internal team chat. Tell him I’m good with deleting the non-prod runner now and adding the narrow non-prod egress guardrail alert. Also confirm we’re treating this as non-prod only, with production not involved, and not turning it into architecture or spend-plan work unless the next cost report still shows leakage.

Sarah at Founders Fund came back with the September offsite logistics ask. Please reply in the existing thread, keep Sarah Kim cc’d, and use my short in-thread signoff. Set the title as Q3 product/operating discussion. Attendees are me, with Devon optional if they want product context. Keep it to a 25-minute slot, no data-room appendix, and if slides are needed they should be lightweight product context only. Make the boundary explicit: this is not a fundraising deck, company-readiness review, or B-round restart.

Blueline followed up after the repair window with an optional valve replacement estimate. Please reply on the repair thread, cc Sarah Kim, and use my short in-thread signoff: decline the optional valve replacement for now and ask them to send the final receipt for the completed leak repair only. Then text Jamie that the leak repair is done and I’m not approving extra work this week.

Pricing check with Devon is done. Create a new internal decision record and quote-guidance doc titled “Mercury pricing decision — hybrid platform minimums.” Record that we’re no longer treating Mercury pricing as a pure-MAU-versus-pure-seat debate. After looking at the first design-partner usage patterns, new Mercury design-partner and post-launch quotes use the hybrid model: - Pilot platform minimum: $2,500/month, including up to 50 monthly active developers - Growth platform minimum: $7,500/month, including up to 200 monthly active developers - Overage: $1,000 per additional 50 monthly active developers - Enterprise add-ons like SSO, audit logs, and advanced admin controls are quoted separately only after those features actually ship Add the rationale plainly: this keeps the usage-ramp logic honest while giving Devon a cleaner floor for forecasting and procurement conversations.

HR needs the final retreat catering count lock. From: HR To: Morgan Chen Cc: Sarah Kim <sarah@atlas-test.com> Date: Tue, 29 Aug 2023 09:12 AM Subject: Retreat catering count lock — due today Hi Morgan, We need the final lunch count lock today. Tava Kitchen can hold the order at the following counts if we confirm by 2:00 p.m.: - Total attendees: 33 - Standard meals: 23 - Vegetarian: 5 - Vegan: 3 - Gluten-free: 2 Optional add-on: afternoon snack service is $420 before tax and service if the team wants it added. Please reply on this thread with a yes/no on the lunch counts by 2:00 p.m. HR will keep the budget line on our side. Thanks, HR Please reply on the count-lock email thread and cc Sarah Kim. Confirm the Tava Kitchen lunch counts exactly as listed: total 33, 23 standard, 5 vegetarian, 3 vegan, and 2 gluten-free. Keep HR as the budget owner, and decline the afternoon snack service or any dinner/snack add-ons unless HR provides a separate approved budget path.

Kara sent the cleaned Mercury landing-page copy. From: Kara <kara@kestrel-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Tue, 29 Aug 2023 11:06 AM Subject: Re: Mercury landing page copy Hi Morgan, Thanks for the earlier guardrails. I cleaned the section up and removed the enterprise-proof / login-admin language from the prior pass. Revised section: Headline: Mercury is open to its first design partners Body: Scaffold is opening Mercury to selected teams for live design-partner evaluation. No SSO, audit, admin-control, procurement, or enterprise-validation claims remain in this version. If this wording looks right, can Kestrel move it into final layout on our side as an unpublished draft? Thanks, Kara Please reply by email to Kara at Kestrel, cc Sarah Kim, and sign it from me in-thread. Approve this wording for design layout only, keep the page unpublished until Scaffold gives final publish approval, and thank her for removing the enterprise-proof and SSO/admin claims.

New Foundry invoice came in. From: billing@foundry-test.com To: Morgan Chen <morgan@atlas-test.com> Date: Wed, 30 Aug 2023 08:17 AM Subject: Invoice INV-F-0314 from Foundry Design — $5,950.00 due 2023-09-15 Hi Morgan, Attached is invoice INV-F-0314 for August design implementation support. Invoice number: INV-F-0314 Vendor: Foundry Design Description: August design implementation support Amount due: $5,950.00 Due date: 2023-09-15 Please let us know if you need anything else. Thanks, Foundry Design Billing Attachment: INV-F-0314.pdf Please flag invoice INV-F-0314 from Foundry Design for manual review. Include the vendor, the $5,950.00 amount, the due date 2023-09-15, and the reason: it is above the current AP review threshold. Do not pay it and don’t email Foundry.

Sarah Kim needs budgetary wording she can use with Evergreen without turning it into a procurement packet. Please email Sarah a forwardable blurb using the new Mercury pricing model: Pilot is $2,500/month up to 50 monthly active developers, Growth is $7,500/month up to 200, overage is $1,000 per additional 50 monthly active developers, and SSO/audit/admin add-ons are not quoted or dated until they ship. Make clear this is budgetary guidance only — not a custom roadmap, not a procurement packet, and not a full commercial commitment.

Mom is asking whether Maya should plan to come by over Labor Day weekend. Please send Mom a warm SMS: Jamie and I are keeping the weekend loose, a short Sunday call is better than making plans now, Kibo is fine, and I’ll tell Maya directly if a visit starts to look realistic.

Pricing/counting ambiguity is popping up after Sarah’s first budgetary wording. Discord — #eng-team Thread: Mercury pricing count / preview-only question Date: 2023-08-31 09:08 AM — Jake Quick pricing/counting question from the new Mercury budgetary wording. If an invited Evergreen user clicks in via magic link, looks around the preview/sample data, but never connects a real source or gets to a live sync, is that person part of the “monthly active developer” band for pricing, or not? Trying to make sure we don’t accidentally recreate seat logic by counting everyone who merely lands in the product. 09:12 AM — Anna Martinez I’d keep that separate. Preview/sample-only was deliberately non-activating in the external wave, and activation is still the corrected definition: real_source_connected or first_live_sync_completed within 7 days. If we let preview traffic do double duty as both “kind of active” and “kind of activated,” our reporting gets muddy fast. 09:14 AM — Jake Yep. I don’t want docs or dashboards to imply “opened preview once = active developer.” We also have internal Scaffold folks hitting the invite flow while testing org setup / retry paths. Those definitely shouldn’t roll into anything customer-facing. 09:17 AM — Anna Martinez Agreed on internal/test users staying out. Otherwise we’d be charging off our own dogfood and also polluting the early wave data. 09:21 AM — Sarah Kim I need a customer-safe sentence here because Evergreen may ask why the budgetary range is tied to monthly active developers instead of total seats in their directory. I can say “budget is based on actual monthly active developer usage in Mercury, not provisioned seats,” but I do not want to ad-lib what counts as active if the next question is “what about someone who only opened the preview?” 09:25 AM — Jake That’s exactly the edge case. Right now an invited person can successfully get into the bounded preview experience before anyone has connected a real source. That makes sense for product, but it leaves a counting gray area if someone reads “usage” too loosely. 09:29 AM — Devon Hayes The rule needs to stay forecastable. If this turns into a one-off exception list by customer, finance and budgeting conversations get messy immediately. Broadly: - pricing bands should map to monthly active developer usage, not directory size or purchased seats - internal Scaffold/test traffic should not count - preview/sample-only behavior should not get smuggled into activation reporting What I do not want is an Evergreen-specific carveout or a shadow seat count hiding underneath the usage model. 09:34 AM — Sarah Kim +1 on no Evergreen carveout. If they ask, I need to keep it budgetary and non-custom. I don’t want to imply a procurement packet, special enterprise policy, or roadmap trade in a one-line answer. 09:36 AM — Anna Martinez Also please don’t collapse “preview-only” into “active” just because we can see session events. We intentionally kept sample data preview-only and non-activating. Reporting-wise, that should remain a distinct bucket from actual activation. Otherwise every retention/activation chart gets re-litigated again. 09:41 AM — Jake Understood. My fear was that someone sees “Pilot includes up to 50 monthly active developers” and interprets that as “count every invited person who clicked once.” That gets us right back to seat-ish behavior, just with different words. 09:45 AM — Devon Hayes Right, and the model only works if the number is explainable without a long footnote. “Actual monthly active developer usage” is good. “Actual usage, except for these seven enterprise edge cases” is bad. Let’s not reopen the seat-versus-usage argument every time a broad directory exists. 09:49 AM — Sarah Kim For Evergreen-facing language, I’ll stay with budgetary guidance tied to actual monthly active developer usage and avoid promising how any future enterprise/security packaging would be counted. If they push past that, I’ll pause rather than improvise. 09:52 AM — Anna Martinez Thanks. And on the internal side, let’s keep preview/sample-only separate from the corrected activation definition in every doc/dashboard. Even if pricing uses an activity concept, activation still means real_source_connected or first_live_sync_completed within 7 days. 09:56 AM — Jake Makes sense. I’ll hold off on answering the eng docs question until we have one clean sentence everyone can reuse. Post a short channel note in #eng-team so this doesn’t stay buried in the thread. Clarify that Mercury pricing bands are based on monthly active developer usage, not purchased seats or directory size; Scaffold/internal test users don’t count; preview/sample-only activity stays separate from activation reporting; any Evergreen-facing answer stays budgetary and non-custom; and nobody should reopen the MAU-versus-seat debate or invent enterprise exceptions in customer emails.

Jamie just texted that she’s running late and wants dinner handled. Please place the Friday Lemongrass order: pad see ew for Jamie and my current saved usual order for me. Add a delivery note asking for around 7:30 if that’s available.

Founders Fund needs the printable logistics for the offsite segment. From: Sarah at Founders Fund <sarah-ff@foundersfund-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Tue, Sep 5, 2023 at 9:14 AM PT Subject: Re: September offsite — quick logistics for your segment Hi Morgan, We’re locking the printed agenda and moderator notes for the September offsite. For your 25-minute product/operating slot, can you send over four quick items? 1. The exact segment title you want us to print 2. The two discussion topics you want the moderator to tee up 3. Whether Devon should be listed as an attendee or kept optional 4. Whether Scaffold plans to send any slides or an appendix by Friday, Sep 8 If you are sending materials, lightweight is totally fine — we just need to know whether to hold packet space. Thanks, Sarah Please reply in the existing email thread to Sarah at Founders Fund, keep Sarah Kim cc’d, and use my short in-thread signoff. Title should be “Q3 product/operating discussion.” The two moderator topics are Mercury design-partner learnings and operating metrics. Keep Devon optional for product context. Say we’re not sending a data-room appendix or fundraising materials; if they need slides, it’ll only be lightweight product context. No B-round restart framing.

Kara added two layout elements that need trimming before this goes anywhere. From: Kara <kara@kestrel-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Tue, Sep 5, 2023 at 2:37 PM PT Subject: Re: Mercury landing page — final layout proof Hi Morgan, Attached the latest full-layout proof for the Mercury section. The approved headline and body copy are now in place. I made two layout-only additions to balance the section visually: - a small pill above the body copy that reads "enterprise-grade evaluation" - a blank customer-logo strip under the proof block labeled "design partner logos" for spacing, in case you want to drop marks in later Nothing has been published. If this looks okay, can we move it into staging for your final review? Thanks, Kara Reply by email to Kara, cc Sarah Kim, and sign it from me in-thread. Kestrel can move this into unpublished staging for final review only after removing the “enterprise-grade evaluation” pill and the design-partner logo placeholder. Do not approve public publishing, and don’t add roadmap detail or proof-claim language.

The AWS follow-up looks contained now. From: AWS Budgets To: Morgan Chen Date: Tue, Sep 5, 2023 07:08 AM PT Subject: Scaffold weekly budget report — Sep 5 Account: Scaffold Budget period: Sep 1, 2023 – Sep 30, 2023 Budget: $15,000.00 Forecast end-of-month spend: $15,403.12 Forecast variance: +$403.12 (+2.7%) Current month actual as of Sep 5 06:45 PT: $2,186.44 Top variance drivers - us-west-2 non-prod NAT Gateway egress: projected $564.18 for September; current daily run rate is back near baseline and materially below the Aug 24–31 spike window. - Compute, storage, and data transfer in production: within normal weekly bands. - No new sustained egress growth detected in the non-prod account after cleanup. Alarm correlation - Production alarms associated with prior NAT egress anomaly: 0 - Customer-facing incidents associated with prior NAT egress anomaly: 0 - Active budget alarms in production networking categories: 0 Notes This is a weekly budget summary generated from current month actuals and forecasted run rates. The account remains slightly above budget on projection, but the previously elevated us-west-2 non-prod NAT Gateway egress is no longer trending as an active spike. Send Marcus a short DM in the current internal team chat: the non-prod egress spike looks contained, keep the narrow guardrail alert, and don’t turn this into architecture work or spend-plan work unless the next report regresses.

Please send HR an internal email and cc Sarah Kim for the retreat lunch day-of details. Sarah Kim should be the on-site delivery contact. Confirm the Tava Kitchen delivery window as 11:45 a.m.–12:15 p.m.; keep the locked counts unchanged at 33 total: 23 standard, 5 vegetarian, 3 vegan, and 2 gluten-free. Use plain dietary labels only, and decline printed menus or any extra add-on.

Devon and I just went through the board/H2 implications of Anna’s first September Mercury package. I want the useful parts preserved without letting this turn into outside proof. September Mercury metrics package Prepared by: Anna Martinez Date: Sep 6, 2023 Audience: Morgan, Devon, Jake Status: Internal working package Data cutoff: Sep 4, 2023 EOD PT Purpose Separate the Mercury reporting into two layers: 1) a board-operating layer built on the sturdier cohort view, for leadership/board operating alignment 2) an internal weekly diagnostics layer for operating decisions only Read-this-first - The cleaner story is real: September design-partner activation is materially better than the repaired May baseline when measured on the current 7-day activation definition. - Admin setup friction is lower across the design-partner cohort. - Evergreen Bank now gives us real enterprise-pattern evidence beyond preview-only usage. - The uncomfortable part is still true: expansion is uneven account to account, and the weekly activation movement is still internal-only. - Recommendation: use the board-operating layer for board/H2 prep context; do not treat this package as external-ready proof. Board-operating layer Definition used here Activated within 7 days = account hit either real_source_connected or first_live_sync_completed within 7 days of workspace creation. Sample-preview-only accounts are excluded from this denominator. 1) Comparable activation view | Cohort / cut | Eligible accounts | Activated within 7 days | Rate | Notes | | --- | ---: | ---: | ---: | --- | | Repaired May comparable design-partner cut | 10 | 4 | 40% | Corrected May definition; useful baseline only, not for patching older external charts | | Current design-partner cohort (workspaces created Jul 31–Aug 27) | 11 | 8 | 73% | Current Mercury test flow with real-source/live-sync evidence | 2) Admin-friction view | Metric | Repaired May comparable cut | Current design-partner cohort | Direction | | --- | ---: | ---: | --- | | Median days from workspace creation to first invited member accepted | 5.4 | 2.3 | better | | Accounts needing manual admin intervention after initial setup | 6 / 10 | 2 / 11 | better | | Accounts reaching first non-owner user within 7 days | 3 / 10 | 7 / 11 | better | 3) Evergreen Bank evidence row - Evergreen is the first account in this set that looks like a real admin + user handoff rather than a single champion walking preview data. - Through Sep 4, Evergreen shows 4 admin users touching setup/settings, 12 invited members, 3 real sources connected, and successful live-sync activity on 8 distinct days. - The account has repeated usage after setup rather than a one-day spike. - This is meaningful enterprise-pattern evidence for internal/board operating context. - It is still one account, not a blanket enterprise-validation claim. 4) Board-operating takeaways - Activation is cleaner than May on the sturdier definition. - The setup/admin handoff is materially less fragile than it was in the May read. - Evergreen adds substance to the enterprise story because usage includes real-source and live-sync behavior, not just sample preview clicks. - This supports board operating alignment and selective H2 prep. - It does not prove retention is solved, prove expansion, or make the story external-ready. Expansion / retention caveats 30-day account-level read across active design-partner accounts | Account | Early pattern | Read | | --- | --- | --- | | Evergreen Bank | invited-member count and weekly active usage both up from initial setup week | best positive signal in set | | DP-02 | flat after first successful sync | no expansion evidence yet | | DP-03 | second admin onboarded, user activity otherwise flat | mild improvement only | | DP-04 | bursty usage after setup, then quiet | uneven | | DP-05 | preview-heavy, later real-source connection but low repeat activity | still fragile | | DP-06 | steady low-volume usage | flat | | DP-07 | small increase in active users | positive but small | | DP-08 | decline after initial champion activity | negative / unclear | Roll-up - 3 accounts look meaningfully up - 3 look mostly flat - 2 remain choppy or down - Net: expansion is uneven and not ready for a solved-story claim Internal weekly diagnostics layer — internal only Do not use the weekly movement below in board notes, board charts, investor materials, or external/customer communication. This layer is for operating decisions only. | Workspace-creation week | Eligible accounts with full 7-day window elapsed | Activated within 7 days | 7-day rate | Sample-preview-only exclusions | Notes | | --- | ---: | ---: | ---: | ---: | --- | | Jul 31 week | 3 | 2 | 67% | 1 | first cleaner post-preview split | | Aug 7 week | 3 | 2 | 67% | 1 | invited-member handoff improved | | Aug 14 week | 3 | 2 | 67% | 0 | steady but still small-n | | Aug 21 week | 2 | 2 | 100% | 1 | strongest week so far | | Aug 28 week | 0 | pending | pending | 0 | full 7-day window not complete at cutoff | Internal notes on the weekly layer - The direction is useful for operating decisions and aligns with what Jake is seeing qualitatively. - The sample sizes are still too small and too sensitive to account mix for external use. - The weekly cut is helpful for debugging admin/setup friction and sample-preview exclusions. - Keep it explicitly internal-only. Suggested framing for Morgan / Devon - Safe to say the September package is materially better than the May operating read on cleaner activation and admin handoff. - Safe to say Evergreen contributes meaningful enterprise-pattern evidence. - Important to keep the downside in the same sentence: expansion remains uneven, and the weekly activation movement is not for board/external charts. - Avoid any statement that implies retention is proven, expansion is solved, or the package is ready as outside proof. Create and save an internal prep document titled “September Mercury metrics package — board and H2 prep bounds,” then send an internal email to Devon and Anna saying it’s ready. The doc should keep the two-layer reporting split: board-operating alignment and selective H2 prep are supported because real-source/live-sync activation is cleaner than May, admin friction is down for the design-partner cohort, and Evergreen adds enterprise substance. It also needs to state plainly that expansion is still uneven, weekly activation movement is internal-only, and this package does not prove retention, solve expansion, or make the story external-ready.

Sarah’s instinct to check this before replying is right. From: Sarah Kim <sarah@atlas-test.com> To: Morgan Chen <morgan@atlas-test.com> Date: Wed, Sep 6, 2023 at 11:18 AM PT Subject: Evergreen follow-up — safe wording on early usage? Hi Morgan, I saw Anna’s internal Mercury package this morning and the Evergreen rows look encouraging. Evergreen asked generally whether we’re seeing useful signal from their first stretch of usage and whether we have anything we can share back yet. Before I reply: is it okay to say their first usage is "already showing enterprise validation," or is that too strong? Also, should I include any chart or short usage summary in the next note, or keep it to qualitative thanks / we’re learning from the eval? I know we don’t want to drift into roadmap or procurement language — just want the safest line here before I send anything. Sarah Draft an internal email reply to Sarah Kim. Say she can thank Evergreen for useful early usage and frame it as input that helps us improve the design-partner evaluation. Do not send the activation chart or the internal weekly movement, and don’t make retention, expansion, procurement, SSO, roadmap, or enterprise-validation claims.

AP finished the Foundry review. From: Devon Hayes <devon@atlas-test.com> To: Morgan Chen <morgan@atlas-test.com> Date: 2023-09-06 08:41 -0700 Subject: Fwd: Foundry INV-F-0314 manual review Morgan — Forwarding the AP review summary on Foundry. This is INV-F-0314 for August design implementation support, $5,950.00 due 2023-09-15. Finance had it on hold only because it tripped the manual-review threshold. They checked the attached August timesheet and it ties out. Can you sanity-check that we’re okay to release it for the reviewed August support work only, not as blanket approval for any new Foundry scope? — Devon ---------- Forwarded review note ---------- Vendor: Foundry Invoice: INV-F-0314 Amount: $5,950.00 Due date: 2023-09-15 Description: August design implementation support Reason for hold - Invoice amount exceeded the manual-review threshold and was routed to AP review before disbursement. Manual review checks - Attached August timesheet total matches billed amount: 34.0 hours x $175.00/hr = $5,950.00 - Service dates on the timesheet fall within August and map to design implementation support hours - No duplicate invoice for the same vendor / period / amount found in the AP queue - No arithmetic discrepancy or rate variance noted Review result - Manual review completed with no exception beyond the threshold-based hold Current status - Payment remains on hold pending internal release after review Please reply internally to Devon: INV-F-0314 can be released after manual review for the reviewed August design implementation support only. Make clear this is not approval for any new Foundry scope.

Devon’s draft is close, but the enterprise-package line will create exactly the wrong expectation. From: Devon Hayes <devon@atlas-test.com> To: Morgan Chen <morgan@atlas-test.com> Date: 2023-09-07 09:18 -0700 Subject: Evergreen budgetary quote draft Morgan — First pass at the Evergreen budgetary language is below. I kept the platform bands aligned to the Mercury pricing sheet, but I left a placeholder on the enterprise package since I assume procurement will ask. If this looks okay, can Sarah use it externally? — Devon Draft: “For Evergreen’s initial Mercury rollout, current budgetary pricing is: Pilot — $2,500/month platform minimum, including up to 50 monthly active developers. Growth — $7,500/month platform minimum, including up to 200 monthly active developers. Overage — $1,000/month for each additional 50 monthly active developers. Enterprise SSO/audit/admin package TBD for procurement. We can treat this as a preliminary quote for budgeting and procurement planning, and tighten language once the enterprise package is defined.” Please redline this and email Devon internally. Keep the Pilot and Growth platform-minimum bands as budgetary guidance. Remove or hold back the unshipped SSO/audit/admin add-on language, avoid a procurement-packet tone, and don’t create Evergreen-specific exceptions or reopen the pricing-counting rule.

Pinecone’s revised wording is still too broad. From: Rishi Patel <rishi@atlas-test.com> To: Morgan Chen <morgan@atlas-test.com> Date: 2023-09-07 11:06 -0700 Subject: Fwd: Pinecone troubleshooting wording Morgan — Pinecone sent back revised troubleshooting copy and wants a yes/no from us. The line below is the part I’m not sure about. Can I tell them this is approved? — Rishi ---------- Forwarded draft excerpt ---------- Troubleshooting signature verification If requests are failing verification after passing through a proxy or connector layer, Scaffold normalizes connector headers automatically, so signature casing and alternate signature locations are accepted. Required headers - `X-Scaffold-Workspace` - `X-Scaffold-Signature` Email Rishi with a narrower edit he can send back. It’s okay to mention the signature-header casing fix and the header-only contract, but remove any claim that Scaffold normalizes all connector headers or supports alternate signature locations.

Mom is asking again about Labor Day spillover plans and whether Maya should try for a Saturday visit. Please send Mom a warm SMS: this weekend still needs to stay loose, I’ll call Sunday afternoon, Kibo is fine, and I’ll coordinate directly with Maya if a short visit becomes realistic. Keep work out of it.