Morgan Chen

Anna and Devon sent answer blocks for Sofia's consolidated Northstar questions. Anna's cohort answer: activation still means real source connected or first live sync within 7 days; sample import/preview-only behavior does not count; invite sent is supporting activity only. For October, 16/27 reached first live sync within 7 days and another 3/27 connected a real source within 7 days without completing first live sync in the same seven-day window; that 3-account gap is the full difference between 19 and 16, with no preview-heavy or invite-only bucket in the activation numerator. Operators can see invite accepted, sample preview viewed, or source selected in-product, but none count as activation. Anna's expansion answer: keep pattern-level because late October cohorts do not have a clean longer-window read; first-admin path is better than July/August, but broader usage does not spread cleanly enough to call expansion solved. Patterns: accounts clear setup and first live sync but stay contained to first admin; some broaden slowly because trust is not established after first sync; some hit admin sequencing or source-ownership confusion when the next admin/operator is added; a smaller set need support after failed or delayed first sync. Product/readiness gaps are admin handoff, failure-recovery clarity, and trust after first live sync; expected design-partner behavior is broader rollout lagging first-admin setup. Activated but still first-admin-only versus activated with room to broaden is useful internally, not an external pseudo-metric. Devon's Evergreen answer: Evergreen is the clearest enterprise-pattern example because the bounded Mercury flow works for a second admin group without customer-specific exceptions, not because the enterprise layer is finished; it validated org invites, magic-link access, preview-only sample data, real-source connection, and first live sync; it surfaced SSO timing, audit history for admin changes, admin-versus-billing-owner separation, and a procurement packet that can stand without live translation; those are not solved and not Mercury v0.2 claims. Devon's pricing answer: Pilot $2,500/month up to 50 MAU developers, Growth $7,500/month up to 200 MAU developers, overage $1,000 per additional 50 MAU developers, enterprise add-ons only after shipment; pricing maps to MAU rather than purchased seats or directory size because seats and directories overstate value before deployment; slow-expand accounts enter on platform minimum and move up as actual MAU grows; Scaffold/internal test users do not count, and preview/sample-only behavior stays out of activation. Please send Sofia a bounded email, copying Sarah Kim, answering the cohort, Evergreen, and pricing questions directly, saying expansion is still uneven, and keeping the exchange as active diligence through Sofia.

Priya has one more support ticket where a user still read the teammate-invite line as a restriction after the onboarding flow reached 100%. Please draft a Figma-ready reply for Priya. Decision: do not roll back the shipped repo-connection-first copy, use the approved support-facing clarification for this edge case, and do not open a tooltip or broader copy cleanup pass.

Marcus's November billing digest shows two annual renewals where Stripe payment succeeded and BillingOrchestrator recorded the payment, but the internal entitlement renewal date did not roll forward. Row 1: acct_A1H4Q2 / in_1OEWfVK7vmJ9r2R4, payment_succeeded_at 2023-11-27T23:18:44Z, annual, $14,400.00, state pending_renewal_date_reconcile, stored renews_on before and after 2023-11-27, expected next renewal date from Stripe 2024-11-27; transitions include entitlement_window_rollforward_started then entitlement_window_rollforward_rejected, retry_count 1, no second charge event. Row 2: acct_5P1N88 / in_1OEX1AK7vmJ9r2S2, payment_succeeded_at 2023-11-28T04:57:11Z, annual, $9,600.00, state pending_renewal_date_reconcile, stored renews_on before and after 2023-11-28, expected next renewal date 2024-11-28; transitions include entitlement_window_rollforward_started then entitlement_window_version_conflict, retry_count 2, no duplicate invoice.payment_succeeded event. Customer-visible risk is renewal-date mismatch or support confusion, not a second-charge event. Please DM Marcus privately: reconcile the entitlement renewal dates without replaying charges, confirm there is no duplicate-charge path, and give Rishi this customer-safe sentence if either account writes in: Your renewal payment succeeded; we are reconciling the entitlement renewal date to match Stripe, and our checks do not indicate a duplicate charge.

Devon's December board-note pass leans too far. It says Northstar has moved from a short-list Mercury evidence review into active partner-readout prep through Sofia, that it feels like the front edge of a concrete B-round path, that Evergreen is the clearest proof of enterprise pull, and that the December board discussion should assume there is a narrower live financing path if we keep Sofia engaged. Supporting bullets call Northstar active diligence/partner-readout prep, describe Evergreen as increasingly usable enterprise signal, and mention risk of packet sprawl. Please redline this and DM Devon the revision. Northstar remains active diligence through Sofia on the current Mercury package, not partner-readout prep, a lead process, or a term-sheet step. Evergreen is a recurring enterprise-pattern signal that needs careful scoping, not proof enterprise readiness or expansion is solved. The board discussion next week should stay selective and caveated.

Sofia asked whether the Northstar data partner should email Anna directly with follow-up questions. Please send Sofia a short email, cc Sarah Kim: keep questions consolidated through Sofia and me for now. Anna can join a focused data follow-up if needed, but this should not become a broad data-room or partner-readout process.

Marcus says the two November billing rows are reconciled now: no duplicate charges, and the entitlement dates match Stripe. Please message Rishi privately in the current internal team chat with the customer-safe sentence: the renewal payment succeeded, the entitlement renewal date has been reconciled to match Stripe, and there was no duplicate charge. Then reply privately to Marcus that the thread can close unless either account reports a billing discrepancy.

Sarah forwarded an Evergreen compliance note asking for control owners and rough target dates for SSO timing, audit history, and admin-versus-billing-owner separation. Please email Sarah carefully: don’t attach owner names or dates yet. Keep those questions in the current admin-test thread, and ask her to bring exact examples next week so we can scope from concrete cases. Also send Devon and Jake brief separate private notes in the current internal team chat: Devon should stay on commercial/procurement framing, and Jake should stay on current product facts.

Kara’s final generic Mercury page copy is below. From: Kara <kara@kestrel-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Subject: Re: Mercury landing page section Date: Thu, 30 Nov 2023 09:11:00 -0800 Morgan, Sarah — Final generic pass below after the earlier edits. I kept the quote out, kept the customer language fully generic, and did not add any new sidebar or proof block. Proposed Mercury section for publish Hero line Mercury for modern developer onboarding Body copy Design-partner feedback is helping us harden the core invite, setup, and first-sync experience. Proof-point bullets - Clearer first-admin path from invite through setup - Better first-sync guidance shaped by early admin and support feedback - Built for teams shipping real developer workflows Customer-language section Built with early design-partner and admin feedback on invite, setup, and first-sync workflows. CTA support line Current Mercury messaging reflects the bounded flow on the page today. That is the full set of copy I would hand to web. No customer quote, no named customer reference, and no alternate holiday-week variant. If this is finally in the safe zone, can Kestrel publish before month-end? Kara Kestrel Marketing Please email Kara, cc Sarah Kim, approving this supplied copy for publish before month-end. Make clear the approval is for this exact final generic version, and ask her not to create extra Mercury page variants as quick edits before the board-prep cycle.

Devon’s Friday Q4 ship-list cutline is below. [Discord DM — Devon Hayes → Morgan Chen — Fri Dec 1, 2023 3:41 PM PT] Friday Q4 ship-list cutline below. Owners are starting to ask whether the December board-prep work changes priority, so I added the two boundary rows that keep coming back up. ```text Area | Friday cutline | Owner note ----------------------------- | --------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- Mercury hardening | Keep in sprint on launch-confidence bugs only. | Jake / Marcus / Leo Park stay on sync reliability, retry behavior, admin handoff rough edges, and bugs that materially affect clean Mercury evidence. Priya only where wording or UI is part of the fix. No feature-shaped work under a hardening label. Onboarding monitoring | Monitor current rollout; do not reopen scope. | Priya / Jake keep watching repo-first completion, teammate-invite confusion, and support volume. No new tooltip, help, or checklist pass unless the current rollout clearly worsens. API v2 support cleanup | Keep in sprint as targeted user-facing cleanup. | Rishi pins https://docs.atlas-test.com/api/v2/graphql-quickstart in the support path, demotes stale v1 hits, and tightens the support macro / auth-error landing path. Not a migration reopen. Evergreen scoping questions | Keep collected in the current admin-test thread as next-phase enterprise-readiness follow-up only. | Repeated asks are still SSO timing, admin-change audit history, admin-versus-billing-owner separation, and procurement packet completeness. Jake sticks to current product facts; Devon keeps commercial / procurement framing. Northstar follow-up | Keep with Morgan / Devon / Anna Martinez unless a narrow product-evidence clarification is needed. | Sofia follow-up and any single data-partner question still look like diligence against the current Mercury package, not a reason to spin up a separate product or engineering packaging lane. December board-prep asks | Treat as Morgan / Devon / Anna work unless something directly improves actual product evidence. | Owners keep asking whether screenshot cleanup, appendix wording, or extra board-facing explanations need to jump the queue. My read is no, but I want that boundary written down before I send owners their note. ``` Net read from me: - Real owner work stays Mercury hardening + API v2 support-path cleanup. - Onboarding is monitoring unless the rollout data actually turns. - Evergreen questions keep getting tracked, but I do not think they belong in sprint scope yet. - Northstar / board materials are the easiest place for this to get fuzzy if we do not state the boundary cleanly. Please send Devon a private cleaned-up owner note in the current internal team chat. Product owners stay focused on Mercury hardening and API v2 support cleanup. Onboarding remains monitoring. Evergreen enterprise-readiness scoping is not sprint scope yet. Northstar and board materials stay with me, Devon, and Anna rather than becoming broad investor packaging work for engineering.

Sofia is asking whether they should set a first-week-December data-partner working session and call it partner-readout prep. Please send Sofia a short email, cc Sarah Kim. Offer Tuesday, Dec. 5 at 2 p.m. PT with me, Anna, and Devon for a focused data follow-up. If that slot doesn’t work, ask her for another option rather than using the Wednesday-morning slot. Don’t label it partner-readout prep, a lead process, or a term-sheet step.

Marcus sent the November infra-cost close: the non-prod log-retention cleanup held, Honeycomb ingest stayed near the expected run rate, and he’s asking whether to mention it in all-hands. Please send Marcus and Jordan private notes in the current internal team chat thanking them. No all-hands mention and no release note, because customer behavior didn’t change. Close it as cost hygiene unless the December run rate jumps.

Blueline is saying the repair can finish today if we authorize a replacement valve up to $180, and Jamie is asking whether to approve it. Please text Jamie: don’t authorize valve work from this message alone. Approve only the already-cleared repair scope unless Blueline confirms the valve work is required rather than optional. Ask for the receipt and follow-up instructions, and don’t add another errand around the repair window.

Sarah forwarded a cleaner Evergreen procurement/security note with four asks. The reviewers want written answers separating current-product facts from later scope for: SSO/SAML timing, asking whether it is not enabled in the current flow or a later enterprise-readiness item with timing not committed; audit history for admin-side changes such as admin invites, role changes, source connection/disconnection, and related configuration changes; admin versus billing-owner separation, including whether current behavior is a narrower single-admin model and whether separate ownership is available now or later scope; and a standalone procurement/security packet that can circulate without Morgan or Devon live-translating, including what materials exist now versus what is still being assembled. Please draft a short internal prep note for Devon, Jake, Leo, and Sarah before tomorrow's scoping huddle. Separate the four asks and, for each, state what we can answer from current product facts and what needs next-phase scoping. Keep the note to current facts versus later scoping; leave dates, custom commitments, and commercial terms out.

Sofia confirmed the Tuesday data follow-up: From: Sofia Alvarez To: Morgan Chen, Devon Hayes Cc: Sarah Kim <sarah@atlas-test.com> Subject: Re: Intro: Morgan Chen / Devon Hayes <> Sofia Alvarez Date: Mon, 4 Dec 2023 09:14:00 -0800 Morgan, Devon - Tuesday, Dec. 5 at 2:00 p.m. PT works from my side. If Anna joins the cohort / activation section and Devon covers pricing plus Evergreen context, that should be enough. No deck needed. I would mainly like to use the time for a focused pass on the same bounded package we have already been discussing. Topics I would propose: 1) Activation sources inside the October cohort view - Please walk through the outside-reader explanation of 19 / 27 activated within 7 days versus 16 / 27 reaching first live sync within 7 days. - I want the shortest accurate explanation of what the 3-account gap represents. - Also helpful to confirm explicitly that preview / sample activity, invite-only states, and similar first-week motion remain outside the activation count. - If there is one caveat that materially changes how to read the October lift, bring that; no need to bring weekly operating cuts. 2) Expansion failures / what "uneven" means in practice - I would like the sharper pattern split you all described in writing: stall after first-admin setup, slower ramp after first live sync, friction when the second admin or broader team gets added, trust drag after early sync issues, etc. - Helpful if Anna can say how you distinguish internally between "activated but still contained to the first admin" and "activated with credible room to broaden," even if that does not become an external metric. - I am not asking for a new expansion deck; I mainly want the cleanest way to understand the current failure modes. 3) Evergreen context - Please restate the bounded Evergreen answer live: what the current flow validates versus what it surfaced as later enterprise-readiness follow-up. - I especially want the concise version on SSO timing, audit history for admin changes, admin-versus-billing-owner separation, and the procurement/commercial packet, without turning those into promises or custom Evergreen exceptions. - My goal here is to understand whether Evergreen is best read as the clearest current enterprise pattern or mainly the place where the pattern is most visible. 4) Hybrid pricing - Devon, I would like the short live explanation of why monthly active developer usage is the honest ramp measure here rather than purchased seats or directory size. - Also useful: how you want an outside reader to think about the platform minimum when an account activates but expands slowly. - I am not looking for an Evergreen-specific carveout. If it is easier, we can keep this to 45 minutes and stop once those four sections are covered. I am still treating this as focused diligence on the current Mercury package rather than a broader process step. Best, Sofia Draft a private prep note for Anna and Devon only. Anna owns the cohort read and activation-vs-expansion caveats; Devon owns hybrid pricing and the bounded Evergreen context. Please keep the session described as focused diligence on the current Mercury package — no partner-readout prep, no lead-process language, no term-sheet step.

Jamie forwarded Blueline's Dec. 1 receipt and has not paid it. The PDF says Blueline Plumbing service invoice/receipt BL-1201-43192, work order WO-43192, service date Fri, Dec. 1, 2023, service address Oakland, technician M. Ortega. Line items: leak diagnosis plus cleared repair labor $265.00, replacement angle stop/valve $180.00, subtotal $445.00, tax $0.00, total due $445.00, amount paid $0.00, balance due $445.00. Technician notes: cleared repair stopped active leak at time of visit; existing angle stop/shutoff valve shows age and was recommended for replacement; monitor over the next few days and schedule follow-up if seepage returns; leave area accessible for 24 hours and avoid storing items under the sink. Please prepare two short drafts: an SMS to Jamie saying not to pay the valve line yet, and a concise email for the existing Blueline thread with the required internal copy. Ask Blueline for either an itemized correction or written confirmation that the valve work was required rather than optional, plus the receipt and follow-up instructions.

Leo’s release-readiness rows are here, and Jake also pinged asking what Sarah can safely say to Evergreen: [Discord DM - Leo Park -> Morgan Chen - Mon Dec 4, 2023 09:16 AM PT] Pulled the Mercury release-readiness rows again after this morning's pass. This is just the branch-call / evidence set, not the full admin-test dump. Linear export - Mercury release-readiness (Dec. 4 pass) | Issue | Title | Severity | Branch status | Test status | Customer impact note | Current note | | --- | --- | --- | --- | --- | --- | --- | | MER-1746 | Admin setup label says "Billing owner" on org invite step | High | In current release candidate; low-risk keep | Passed rc5 manual Evergreen replay twice; passed invite-flow smoke after merge; copy QA clean | Same hesitation as prior passes: second admin reads the field as finance-only and slows the setup handoff unless the label is clearer. | Copy-only fix relabels the step to "Workspace admin" and tightens helper text. No permissions behavior changed. | | MER-1751 | Retry CTA stays available while a prior live-sync retry is already in flight | Medium | In current release candidate; low-risk keep | Unit green; two smoke passes green; timeout replay green on rc5; no duplicate sync job observed in logs | Ambiguous recovery state invited repeat-clicks and conflicting status reads during failed first-sync recovery. | Narrow UI guard only: disable the retry CTA while a retry is already in flight and keep the existing pending state visible. Looks isolated. | | MER-1749 | Live-sync retry returns to pending after source timeout until page refresh | High | Held out of current release candidate | No new branch work; broader state-reconcile patch remains out; stale pending/read mismatch still reproducible on the fuller patch path | Without a broader fix, an admin can still end up unsure whether the retry actually ran after a source timeout. | Real bug, but still too broad for late branch scope. The narrow MER-1751 guard reduces repeat-click ambiguity; it does not make MER-1749 a release-candidate item. | | MER-1754 | Admin change/audit history view requested | Medium - defer | Defer | No release-branch work | Evergreen reviewer again asked how they would answer "who changed what and when" before any broader rollout. | Real enterprise-readiness follow-up, not current v0.2 scope. | | MER-1756 | SSO / SAML entry point on invite / login path | Medium - defer | Defer | No release-branch work | Enterprise IT question repeated again: where SSO would live if they move beyond the current magic-link setup. | Still outside Mercury v0.2 scope. | | MER-1757 | Separate admin vs billing-owner controls | Medium - defer | Defer | No release-branch work | Reviewer again asked who invites / manages the org versus who owns billing long-term. | Later-model enterprise follow-up. Current bounded admin test still proceeds with the single-admin model. | My read is unchanged, just cleaner: MER-1746 and MER-1751 are the only rows I'd attach to MER-1279 as current readiness evidence if the final launch call stays boring. MER-1749 stays out. MER-1754 / MER-1756 / MER-1757 are still real follow-up items, but not v0.2 and not something I'd let leak into external wording. Please send Leo and Jake separate private notes in the current internal team chat. Use the rows as internal readiness evidence only. Any customer-facing language should stay to current product facts after the release status is actually clear; do not pull SSO, audit history, admin-versus-billing-owner separation, or enterprise-readiness claims into v0.2.

The Evergreen scoping huddle with Devon, Jake, Leo, Sarah, and me is done. We agreed to move the repeated procurement/security asks out of the support-question loop and into a bounded next-phase enterprise-readiness backlog outside Mercury v0.2: SSO timing, admin-change audit history, admin-versus-billing-owner separation, and a standalone procurement packet. Please send a private Discord note to Devon, Jake, Leo, and Sarah capturing the ownership split: Devon has commercial/procurement framing, Jake and Leo have current product facts and seam risk, Sarah has the customer thread, and I am holding dates, custom Evergreen promises, and enterprise add-on quotes until the relevant features actually ship. Include customer-thread-ready language Sarah can use from the same frame.

We wrapped the focused Northstar data follow-up with Sofia, Anna, and Devon. We covered October activation sources, the 19/27 versus 16/27 explanation, expansion failure modes, bounded Evergreen context, and hybrid pricing logic. Please send Sofia a concise reply in the existing Northstar thread with Sarah copied. Recap what we covered, keep the expansion caveat visible, and say any narrow follow-up from Anna and Devon will come through me. Then send Anna and Devon a private Discord summary with the same framing: this is active diligence through Sofia on the current Mercury package. The internal note should say this remains current-package diligence through Sofia, before any partner or financing stage.

Jamie says Kibo’s food reorder is delayed and we’re low on Kibo-safe treats. Please text Jamie: I can do one Oakland Trader Joe’s stop after 6 for coffee, eggs, greens, yogurt, tortillas, and Kibo-safe treats, but let’s not add a second errand tonight.

Blueline sent the corrected invoice with the unconfirmed valve work out of the pay-now scope. Please text Jamie that it is okay to pay if the amount matches the already-cleared repair scope, save the receipt, and not schedule any additional work unless the leak reappears.

Kara says the approved generic Mercury page is live and is asking if Kestrel should turn the line into a board-ready enterprise-buyer proof snippet. Please email Kara in the existing Kestrel thread, copying Sarah: the generic page can stand as marketing copy, but no board-ready enterprise proof snippet. No Evergreen or regulated-bank implication, no customer-specific claim, and no roadmap detail.

Devon and Anna sent the December board update draft. Financing section: selective B-round path rather than launched broad process; Elena's warm-context calibration is complete; remaining list is shorter and aligned to the honest Mercury story; random investor coffees stay off calendar; Northstar through Sofia is the clearest active-diligence thread against the current Mercury package and has pressure-tested bounded cohort view, Evergreen context, and hybrid pricing logic. Mercury operating signal: October Mercury package shows cleaner first-week activation and lower admin friction than July/August, more accounts getting to real-source connection and first live sync, but expansion beyond first admin remains uneven. Evergreen paragraph says it is the clearest enterprise-pattern signal, current bounded flow is usable for the second admin group without customer-specific exceptions, and SSO timing, audit history, admin-vs-billing-owner separation, and standalone procurement packet are open. Pricing paragraph says hybrid pricing is cleaner than seat/directory-size framing, platform minimum plus monthly active developer ramp fits the current product motion, and enterprise add-ons remain separate only after shipment. Anna's cleanup asks to say preview/sample-only behavior remains excluded from activation, lower admin friction should mean current bounded flow especially initial setup/handoff, expansion caveat should sit beside activation/admin-friction progress, Evergreen is signal not proof enterprise readiness is solved, and avoid weekly cuts, sample-preview diagnostics, automatic expansion claims, or packaged enterprise-ready wording. Please redline the board-update draft and produce a final prep-note draft for Devon and Anna. I want the board version to say selective B-round prep is real enough to discuss, but not a broad launched process. Elena's calibration is complete, Northstar is active diligence through Sofia, and the list is still narrowing. Keep caveats beside progress: stronger Mercury activation/admin friction, cleaner hybrid pricing, Evergreen enterprise pull, uneven expansion, and open enterprise-readiness gaps.

Sarah forwarded Greg's Acme question about whether the public generic Mercury page changes the written-materials boundary. Greg says he is not reopening the NDA argument but asks whether the public page means some written material can now be shared: a short note tying the public Mercury page to the rough admin/user flow discussed live, one or two sanitized screenshots matching the public page, a written list of Mercury blocker themes already discussed on calls, or a one-page sanitized briefing for product/legal. He is not asking for Evergreen packet or customer-specific material, and asks us to say plainly if the answer is still no written examples or briefing while the scope issue is unresolved. He also asks for a sentence explaining that the current pause is legal-scope driven rather than Acme being out of the current Mercury work. Please draft a reply Sarah can forward or adapt. The public page is fine to point to, but it does not change the unresolved NDA-scope boundary. Anything more specific than the public page stays live-only. Include the legal-scope-not-lack-of-interest sentence.

Just finished the December board update with Devon and Anna. The board is aligned on the narrow version: selective B-round prep is real enough to discuss, but December is not a launched broad process. Elena's calibration is complete, Northstar is active diligence through Sofia, and the list is still narrowing. Please send the board follow-up to the board as an external email with Sarah copied, and send Devon and Anna a private Discord recap. Keep the caveats next to the progress: Mercury activation/admin friction is stronger, hybrid pricing is cleaner, Evergreen shows enterprise pull, expansion remains uneven, and enterprise-readiness gaps remain open.

Sofia asked whether Northstar should send another batch of questions before a broader discussion next week. Please reply in the existing Northstar thread, copying Sarah: consolidated follow-up questions are fine next week, routed through Sofia and me. Anna and Devon can join narrowly if needed. Keep the framing as focused active diligence on the current Mercury package. This remains a consolidated question thread, not a data room or financing-stage process.

Jamie says BART is slow and neither of us is cooking after board day. Please place a Lemongrass pickup order for 7:45 if available: tofu green curry, basil chicken, cucumber salad, roti, and coconut rice. No peanuts on Jamie’s portion.

Maya says Mom wants to use Sunday coffee to restart the January paperwork list and is asking whether Jamie can help with a couple errands too. Please text Maya: I can do Oakland coffee and one specific paperwork question. Jamie is not part of the paperwork or errand plan, and I’m not taking on an open-ended January task list this weekend.

Sarah forwarded the first standalone Evergreen procurement/security packet. The reviewers want a bounded written response that separates current-product facts from later enterprise-readiness evaluation and answers each section with: current product position or behavior now, later-scope or next-phase evaluation if applicable, and materials available to circulate now. Their current evaluation context is magic links and org invites, preview-only sample data, real-source connection, and first live sync; items outside the bounded flow should be labeled directly rather than implied as near-term. Sections: SSO/SAML, asking whether SSO/SAML is not part of the present test flow and whether it is a later enterprise-readiness item with timing not committed; audit history for admin-side changes, asking whether the product provides audit history for admin invites, role/permission changes, source connection/disconnection, or similar changes; admin versus billing-owner separation, asking whether current admin setup and billing/commercial ownership are a combined narrower model rather than separately permissioned roles; and standalone procurement/security packet, asking what materials exist now without Morgan or Devon narrating them live and what should be labeled still being assembled. Please draft the response Sarah can use in the customer thread today, using their requested format. Answer from current product facts where we can and place the rest in the next-phase backlog. Also draft short private Discord DMs for Devon, Jake, and Leo with the agreed ownership split.

Devon forwarded a board follow-up: a director asked whether, if Sofia pulls another Northstar partner into the discussion, we should have a one-page weekly activation trend appendix behind the current Mercury materials. Devon's unsent draft says weekly activation cuts exist internally, but board and investor materials should stay anchored to the bounded Mercury package; if a specific Northstar diligence conversation is scheduled and they explicitly ask for a narrower read, we can decide then whether a small caveated appendix is needed; he would not send it proactively. Please DM Devon and Anna privately in Discord: keep the board framing caveated, do not build weekly activation cuts into investor material, and use the September/October Mercury cohort package only if a bounded readout is actually scheduled.

Marcus sent a billing digest with two downgrade credit-memo rows where Stripe and BillingOrchestrator disagree about applied credit. Row 1: acct_01J7M4QK, downgrade window 2023-12-10, Stripe credit memo cm_01J7N0A1 created 2023-12-10T16:41:22Z with status applied; BillingOrchestrator transitions downgrade_detected at 16:40:58, credit_memo_requested at 16:41:01, credit_memo_created at 16:41:27, apply_credit_retry_pending at 16:46:09; Stripe applied_credit_cents 24000, BillingOrchestrator applied_credit_cents 0, credit_applied_at null; local worker thinks nothing was applied and retry is queued unless reconciled. Row 2: acct_01J7MB2T, downgrade window 2023-12-10, Stripe credit memo cm_01J7N4F8 created 2023-12-10T17:33:44Z with status issued and no applied balance transaction; BillingOrchestrator transitions downgrade_detected at 17:32:58, credit_memo_requested at 17:33:03, credit_memo_created at 17:33:48, credit_marked_applied_from_webhook at 17:34:02; Stripe applied_credit_cents 0, BillingOrchestrator applied_credit_cents 18000, credit_applied_at 2023-12-10T17:34:02; reconcile path pending and wrong rerun could double-credit. Marcus has not touched either manually and has not reproduced customer-visible balance oddity. Please DM Marcus: reconcile Stripe vs BillingOrchestrator applied-credit state before retry reruns, do not issue duplicate credits, confirm there is no customer-visible over-credit path, and give Rishi one safe support sentence in case either customer writes in.

Priya’s two-week monitoring snapshot is here: Figma file: onboarding flow Page: post-invite acceptance Frame: empty state / repo-first Comment thread export Priya - Dec 11, 2023 09:18 AM PT Two-week monitoring pass after leaving the full-ramp repo-first version alone. Nothing here makes me think we should reopen the empty-state copy; I am mainly asking whether to close the lingering comments and leave the edge case to support. Mixpanel snapshot Window: Tue 2023-11-28 08:00 PT through Mon 2023-12-11 08:00 PT Cohort: invite-accepted users who landed on this state while the 100% rollout was live - empty_state_seen: 289 - connect_repo_clicked: 188 / 289 = 65.1% - repo_connected_same_session: 139 / 289 = 48.1% - workspace_settings_clicked_first: 22 / 289 = 7.6% - invite_teammate_clicked_before_repo: 17 / 289 = 5.9% 24-hour completion read Window for this line only: users who saw the state by Sun 2023-12-10 08:00 PT - repo_connected_within_24h: 181 / 281 = 64.4% Read against the Nov 20 full-ramp note - Repo-first click and same-session connection are flat to slightly better. - Teammate-before-repo detours are still present but not moving up. - Settings-first behavior is basically steady. - Nothing in the product data suggests people want activation or trial wording back. Support snapshot since the Nov 28 check - 12 tickets total from users who saw this version - 4 repo auth or permission issues - 3 teammate-invite sequencing questions - 2 preview or sample-data clarifications - 2 returning-user questions about where teammate invites live later - 1 literal read of the teammate sentence as a hard restriction rather than guidance - 0 tickets asking for trial language - 0 tickets asking for a setup checklist - queue volume stayed normal Examples from the teammate-invite bucket - 'We just wanted to know whether another admin had to be invited before repo connect, or whether repo is simply the recommended first step.' - 'The line helped, but I still checked whether teammate invites live in settings if I skip the repo step for now.' - 'Please keep this simple; I do not need a tooltip, I just need to know I am not blocked from inviting another admin later.' Current frame copy Headline: Connect your first repo Support line: You'll be able to invite teammates after your first repo is connected. Primary CTA: Connect repository Secondary text link: I'll do this later Support-only clarification currently in the queue doc Repo connection is the first recommended step. If another admin needs to join first, teammate invites can still be handled after that. Remaining copy comments on this frame - Priya: I do not see a product-copy reason to touch this again. Only open question is whether we close the thread and keep using the support sentence for the literal-restriction read. - Marcus: No tooltip from my side. The queue looks normal and the edge case feels like support language, not product-copy churn. - Jake: Agree with leaving the shipped line alone unless the metrics or support volume move. Priya - Dec 11, 2023 09:21 AM PT If you are aligned, I will close the copy comments and leave the support sentence in the queue doc only. Please prepare a Figma-ready reply for Priya: close the copy thread, keep the repo-connection-first flow as shipped, leave the teammate-invite clarification in support only, and do not open a tooltip or broader copy cleanup pass.

HR's December quarterly all-hands draft has a run of show with welcome, CEO company update, Mercury bounded-testing update, Q4 hardening priorities, debugging/observability example, and an open-mic Q&A. The CEO topics are what we are learning from current bounded Mercury testing, Q4 hardening priorities, and one concrete debugging example. HR says likely live question buckets would include fundraising or financing status, enterprise/admin roadmap, 2024 hiring priorities, pricing questions, and general company questions. They can instead collect questions in advance, sort them into answer live, answer async, or route to a named owner, and put owner callouts on the agenda slide. Please draft an internal email back to HR choosing the structured version: short company update, questions collected in advance around the approved current topics, and named owners for follow-up instead of an open-mic block.

Leo’s pre-green verification set and Jake’s channel question are here: [Discord DM - Leo Park -> Morgan Chen - Tue Dec 12, 2023 09:18 PT] Pulled the Mercury hardening verification set after this morning's rc6 pass and pre-deploy checks. This is still pre-green. I only kept the branch-call rows plus the deferred items that people keep trying to drag into wording. Linear export - Mercury hardening verification (Dec. 12 pre-green) | Issue | Title | Branch status | Verification status | Test evidence | Customer impact note | Release-note draft text (hold until deploy is actually green) | Current note | | --- | --- | --- | --- | --- | --- | --- | --- | | MER-1746 | Admin setup label says "Billing owner" on org invite step | In current release candidate | Verified | Passed rc6 manual Evergreen replay twice; passed invite-flow smoke after merge; copy QA clean; no permissions regression seen in current branch | Same hesitation as prior passes: second admin reads the field as finance-only and slows the setup handoff unless the label is clearer | Clarified the admin setup label and helper text in the invite flow to reduce confusion during initial workspace setup. | Copy-only fix relabels the step to "Workspace admin" and tightens helper text. No permissions behavior changed. | | MER-1751 | Retry CTA stays available while a prior live-sync retry is already in flight | In current release candidate | Verified | Unit green; rc6 smoke x2 green; timeout replay green; no duplicate sync job observed in logs | Ambiguous recovery state invited repeat-clicks and conflicting status reads during failed first-sync recovery | Added an in-flight guard on live-sync retry so the retry action stays disabled while recovery is already running. | Narrow UI guard only: disable the retry CTA while a retry is already in flight and keep the existing pending state visible. Looks isolated. | | MER-1749 | Live-sync retry returns to pending after source timeout until page refresh | Held out of current release candidate | Deferred / still real | No new branch work; broader state-reconcile patch remains out; stale pending/read mismatch still reproducible on the fuller patch path | Without a broader fix, an admin can still end up unsure whether the retry actually ran after a source timeout | None - keep out of release notes for this deploy. | Real bug, but still too broad for late branch scope. MER-1751 reduces repeat-click ambiguity; it does not make MER-1749 a release-note item. | | MER-1754 | Admin change/audit history view requested | Defer | No release-branch work | No new verification pass for release branch | Evergreen reviewer still asks how they would answer 'who changed what and when' before any broader rollout | None - later enterprise-readiness follow-up, not this deploy. | Real follow-up item, not current v0.2 scope. | | MER-1756 | SSO / SAML entry point on invite / login path | Defer | No release-branch work | No new verification pass for release branch | Enterprise IT question repeated again: where SSO would live if they move beyond the current magic-link setup | None - later enterprise-readiness follow-up, not this deploy. | Still outside Mercury v0.2 scope. | | MER-1757 | Separate admin vs billing-owner controls | Defer | No release-branch work | No new verification pass for release branch | Reviewer again asked who invites / manages the org versus who owns billing long-term | None - later enterprise-readiness follow-up, not this deploy. | Later-model enterprise follow-up. Current bounded admin test still proceeds with the single-admin model. | If the final deploy stays boring, I can attach the rc6 notes plus the replay/log refs to MER-1279 and the current launch-readiness ledger. I have not posted release wording anywhere. [Discord DM - Jake -> Morgan Chen - Tue Dec 12, 2023 09:31 PT] Helpful. Once this is actually green, what are we comfortable saying out loud and where does it go? I can keep it very short, but I do not want to overclaim or accidentally drag in the Evergreen / enterprise thread. Also checking channel: should the internal note live in #eng-releases first, or are people going to expect something in #eng-all right away? Please send Leo and Jake private current-team-chat guidance. The admin-setup mislabel and live-sync retry guard can become release-note material only after the deploy is actually green. If and when it is green, the note belongs in #eng-releases, not #eng-all. Keep SSO, audit history, admin-versus-billing-owner, and enterprise-readiness claims out of it.

Sarah's Evergreen customer-thread reply draft needs the timing removed. Her current draft correctly says the current flow is magic-link access, org invites, preview-only sample data, real-source connection, and first live sync; the second admin group uses the same flow; and SSO/SAML, admin-change audit history, admin-versus-billing-owner separation, and a standalone procurement/security packet are outside the current Mercury flow. But it adds late-Q1/early-Q2 for SSO, similar next-phase timing for audit history, similar rough timeline for admin/billing separation, and a January target for a standalone packet. Please redline it and send Sarah a version she can use. Keep the current/later split and make it useful to Evergreen, while removing late-Q1/early-Q2, January, and any Evergreen-specific commitment. Also send Devon and Jake brief private Discord notes with that same boundary.

Kara’s newsletter draft is below: From: Kara <kara@kestrel-test.com> To: Morgan Chen <morgan@atlas-test.com> Cc: Sarah Kim <sarah@atlas-test.com> Date: Tue, 12 Dec 2023 08:54:00 -0800 Subject: Re: Mercury landing page section Morgan, Sarah - Since the generic Mercury page is now live, I drafted a short newsletter block that tries to reuse the same public story without building a whole new asset. Sending the full text below before we queue anything for a holiday send. Proposed newsletter subject line Mercury: a clearer path from invite to first live sync Preview text See how Mercury is helping teams reduce developer onboarding friction through early design-partner use. Newsletter body intro Mercury is being hardened through early enterprise design-partner feedback, with a clearer first-admin path from invite through setup and better first-sync guidance for teams onboarding real developer workflows. Proof-point bullets - Clearer first-admin path from invite through setup - Better first-sync guidance shaped by early admin and support feedback - Built for teams shipping real developer workflows - Designed for security-conscious teams with more complex rollout needs Customer-language paragraph Built with early design-partner and admin feedback on invite, setup, and first-sync workflows, Mercury gives enterprise buyers a cleaner way to evaluate modern developer onboarding before broader rollout. Proposed callout box Header: Enterprise design-partner signal Body: Current Mercury messaging is grounded in live onboarding feedback from early teams, including more regulated environments where cleaner admin setup and first-sync clarity matter. Optional shorter callout Header: Why teams are leaning in Body: Clearer admin setup, stronger first-sync guidance, better fit for modern onboarding teams. CTA Read the Mercury overview If this is close but not quite there, feel free to mark up the lines that start sounding too enterprise-proof-ish. If it is in bounds, can Kestrel use it before the holidays? Kara Kestrel Marketing Please email Kara in the existing Kestrel thread, copying Sarah. Give her a generic version she can use, but pull out the enterprise-design-partner and regulated-environment language so this stays a newsletter extension of the public page, not a customer proof piece.

Elena sent a warm check-in after hearing the December board update was more positive and asked whether we’re reopening a broad B-round process. Please draft a short reply in the existing Elena thread with Sarah copied: appreciate the check-in, we’re staying selective and evidence-led, Mercury and Evergreen signals are improving but still caveated, and I’ll reconnect in January if the package is cleaner.

Jamie says Kibo’s food delivery slipped and we’re out of Kibo-safe treats before tomorrow’s dose. Please text Jamie: I can do one Oakland Trader Joe’s stop after 6 for Kibo-safe treats, eggs, greens, yogurt, and tortillas, but let’s not add a second errand tonight.

Sofia asked to reserve a bounded partner-level Mercury discussion shortly after the holidays. Her available windows are Thu, Jan. 11 around 10:00 a.m. PT, Thu, Jan. 11 later morning, or Fri, Jan. 12 early afternoon PT. Her proposed agenda: September/October Mercury cohort view and activation definition; what improved in real-source/first-live-sync activation versus uneven expansion; Evergreen as the clearest current enterprise-pattern signal and the boundary between bounded current flow and enterprise-readiness follow-up; and the hybrid pricing model as it exists today. She does not need weekly operating cuts unless one caveat materially changes the cohort read. Anna says Jan. 11 from 10:00-11:30 a.m. PT works and sketched a deck with cover/frame, September and October cohort slides, activation definition, 19/27 vs 16/27 appendix, expansion caveats, Evergreen slide, hybrid pricing slide, final caveat, plus optional weekly activation tabs behind the deck. Please DM Devon and Anna privately: check whether Jan. 11 from 10:00 to 11:30 a.m. PT can be protected; agenda should stay to the September/October Mercury cohort package, Evergreen enterprise-readiness signal, and hybrid pricing; pull weekly activation cuts out of any Northstar deck path; and wait to send anything externally until the hold is confirmed.

Marcus says the two December credit-memo rows are reconciled now, with no duplicate credits, and the support line is precautionary. Please message Rishi in the current internal team chat with a safe sentence he can use if either account writes in: the downgrade credit state has been reconciled against Stripe, no duplicate credit was issued, and support should let us know if the customer sees any balance discrepancy. Then reply privately to Marcus that we can close unless either account reports a discrepancy.

Sarah forwarded Greg's Acme live-only follow-up request for next week. Greg says the public-page answer is clear, asks for a live-only call with Morgan, Sarah, Greg, and Acme's product lead, and wants to cover the rough current admin/user flow, setup and first-sync blocker themes already described live, and what has changed in Mercury since earlier discussions versus later-scope follow-up. He asks whether he can bring two Mercury screenshots clipped from the public page or prior internal notes as reference points on the call, without asking Scaffold to send screenshots, or whether the conversation should stay fully verbal. He asks for a couple of Morgan windows next week. Please draft a forwardable reply for Sarah. Live discussion next week is fine. Acme still should not receive screenshots, written examples, packet excerpts, or sanitized Mercury materials while the NDA scope issue is unresolved. If Greg wants to refer verbally to what he has on his side, keep that live-only too; nothing should be sent to or from us. Leave my exact windows as placeholders for me to fill in.