marnjito knowR2
Marnji · redesign · r2a — + experts, expert profile, real booking details

Now with the experts directory, expert profiles, and the real booking-details page.

This round keeps the left-shell iteration and folds in Khoa's three notes, each grounded in the current staging build (walked staging.marngi.com — the boxless /chats lane, the /pre-answers swipe stack, and the 20/40-min session model): (1) the answers now flow boxless in the page like GPT/staging, so booking & connect cards can drop straight into the lane; (2) a pre-answers swipe stack — matched experts each draft a take, you swipe through and pick one; (3) the Bookings page becomes real confirmed bookings, not a marketing page. Same warm system.

💬 Boxless lane · cards inline🃏 Swipe stack · pick your mind📅 Real bookings · join / receipts◆ grounded in staging.marngi.com
The screens · clickable prototype (★ = new / changed this round)
★ EXPERTS · DIRECTORY

Experts — the directorymatches staging /experts · the sidebar "Discover" now lands here

Added from staging (Khoa). "Find the right person to ask" — search + category chips + a grid of expert cards (Ask · ★ · tags · From $200). Browsing lives here; the ask flow still goes to the swipe stack.

Open →
★ EXPERT · PROFILE

Expert profile — /experts/{name}photo · rating · 1:1 consultation · Ask / Book · about · topics

Added from staging. The full expert page: verified name + ★5.0 (120 reviews), 1:1 Video Consultation ($200), Ask Chris / Book session, About me, and "can answer questions on the following".

Open →
★ BOOKING · DETAILS

Booking detailsmatches the real bookings page · reached from the Bookings list

Rebuilt to match the real page (Khoa). Who's-joining form + a summary card (duration/date/time/format/payment/Join call/Confirmed/total + expert-payout breakdown) + session-thread link back to the chat + cancellation policy.

Open →
01 · HOME (in the shell)

Home — the ask, inside the left shellsidebar: New question · Discover · Recent conversations · You · profile

The staging-faithful "One question, many minds" ask hero now lives in the main pane; the evolved sidebar carries search/discover, your recent brain conversations, and Book/Connect/Dashboard — the ChatGPT-style shell.

Open →
★ PRE-ANSWERS · SWIPE STACK

Pre-answers — the swipeable perspective stackask → matched experts each draft a take → swipe through → keep talking to one

Khoa's note #2, grounded in staging's /pre-answers. After you ask, a few matched brains each draft their perspective as a stacked deck you swipe through (pills + arrows), each with Ready/Drafting state and a "keep talking to →". This is now the screen the home ask lands on. Interactive — swipe it.

Open →
RESULTS · STACK (matches staging)

Results — the staging pre-answers stackquestion → expert pills → swipe the deck → Continue with X

Rebuilt to match staging (Khoa). Was a vertical list — now the real /pre-answers swipe stack: your question, a row of expert pills, and a deck of cards you swipe (each a Strong-match tag + specialty + the drafted take + "Continue with X"). Interactive — swipe it.

Open →
★ ANSWERS · BOXLESS LANE

Answers — clean lane, Book/Connect on the sideboxless conversation · composer right after the last answer · Book/Connect in a right rail

Khoa's note #1, grounded in staging's /chats. r2a — Khoa feedback: the boxless answer flow is kept, but Book/Connect no longer sit between the last answer and the composer (they blocked "keep chatting"). They move to a persistent right rail, with one subtle contextual "book" nudge only where the brain defers to a human.

Open →
03 · PUBLIC BRAIN

Public brain page — context-aware shellsame shell, logged-out content: Explore + "Create your own brain" + Sign in

A stranger arriving from a shared link sees the same sidebar shell, but it adapts — Explore/discover + a "Create your own brain" CTA + Sign in, not an empty conversations list. Main pane still foregrounds talk / book / connect / sign up.

Open →
04 · ADD TO APP

Add to your app — click-once connectbutton placement + the one-tap consent modal

"Add to Claude / Add to ChatGPT" → a permission modal (the trust moment) → connected. Honest that a fully-silent install isn't possible; consent stays.

Open →
★ BOOKINGS · REAL DETAILS

Bookings — your confirmed sessionsreplaces the old marketing page · the sidebar "Bookings" now lands here

Khoa's note #3. Real confirmed bookings — Upcoming / Past, each with expert, date/time, 20/40-min duration, amount paid, status (Join call / Confirmed / Completed), reschedule, receipt & session notes. Closer to staging's real account.

Open →
BOOK VIA A CONNECTED APP

Booking from another appinitiate in the host → complete & pay on Marnji

The Connect-side booking flow (unchanged): when someone books through a connected brain inside Claude, availability surfaces in the host, they pick a slot, then a handoff brings them to Marnji to confirm-and-pay (Stripe). Reachable from Book buttons.

Open →
06 · SHARE

Share your braincopy-link · social · QR · embed

A dedicated surface to get an expert's brain out there — copy link, social share, a QR, and an embed snippet for their own site.

Open →
07 · DASHBOARD

Owner dashboard — data managementmaterials upload/update · what it knows · who can reach it

Solid, ongoing materials workflow (add/remove sources + re-index), what the brain confidently knows vs defers on, and the three reach doors (Public / Book / Connect).

Open →
Design note

This round — Khoa's three notes (grounded in staging)

1 · Un-boxed the answer lane. I walked staging's /chats/… view: the response isn't in a box, it flows down the page. I matched that — the answers page is now a boxless lane (sticky expert header → conversation in the page → composer), which is what lets us reuse the whole lane to drop booking & connect cards inline between messages, exactly as you wanted.

2 · Added the pre-answers swipe stack. Grounded in staging's /pre-answers/…: matched experts each draft a perspective, presented as a stacked deck you swipe through (expert pills + arrows, Ready/Drafting states), then pick one to keep talking. The home ask now lands here.

3 · Real bookings page. The old Bookings destination was a marketing page; it's now your confirmed sessions — upcoming/past, expert, time, 20/40-min duration, amount paid, Join call / reschedule / receipt / notes — closer to staging's real account. The sidebar "Bookings" now lands here.


Kept from the left-shell iteration: the sidebar shell everywhere (grounded in staging's thin left icon rail, evolved wide + labelled).

Left sidebar shell, everywhere (Khoa's call). R2's top-nav is replaced by a left sidebar on every surface — the convention every AI app now uses (ChatGPT / Claude / Gemini), and a natural fit for a chat-centric product. I grounded it in staging's existing thin left icon rail (toggle · new-chat · discover · help) and evolved it into a wide, labelled sidebar: New question · Search & discover · Recent conversations (brains you've talked to) · Your brains · Book / Connect / Dashboard · profile at the foot.

Context-aware, not two shells. The public brain page (a logged-out stranger from a shared link) keeps the same shell visually, but the content adapts — Explore + "Create your own brain" + Sign in instead of a personal conversations list. One consistent shell, content that knows who's looking.

The missing bridge is in. A new matched-results → expert screen sits between "ask a question" and the answers page: your question → a few best-fit brains (each Talk / Book / Connect) → pick one → talk. "One question, many minds" is preserved and finally has its results moment.

Everything else from R2 is kept — just re-shelled. Talk/Book/Connect on the answers page, the add-to-app click-once + one-tap consent modal, the booking-via-host handoff (initiate in the host → complete & pay on Marnji), the share surface, and the owner dashboard/data-management are all intact inside the new shell. Same warm system (ink #181614, Fraunces + Figtree, cream/terracotta, arch avatars, pill buttons); no end-user MCP jargon.

No end-user jargon. Users never see "MCP" — they see "add to Claude / your apps." The mechanism is invisible; the outcome leads. No Spotify/royalty language anywhere.

Two honest constraints made visible (per your realism note): (1) connecting can't be fully silent — the one-tap consent modal is the trust moment, and tapping "Add" deep-links into the host's own connector flow; (2) booking through an external app initiates in the host but completes and pays on Marnji (Stripe), so money + details stay secure.

Flags for Maya (spec / staging constraints)

  • Connect / add-to-app is concept UI — there's no live connector flow on staging yet; I've designed the button + consent modal to match how the host (Claude/ChatGPT) add-a-connector deep-link would work. Swap in the real host flow when it exists.
  • Booking-via-host assumes availability can be surfaced read-only inside the host and the pay step returns to Marnji — worth confirming the handoff/deep-link is technically viable with Ben before build.
  • Casting/copy is tunable — Lauren Okafor / Dr. Maya Reed are placeholder brains; Chris March is the real staging profile. Real creator photos slot into the arch avatars (I used warm gradient placeholders, no fabricated faces).
  • Public @handle page assumes each brain gets a public route (marnji.com/@handle) — confirm the routing + the "grounded, cites, honest" claims are safe to state publicly.
  • Sidebar behaviours to confirm at build: collapse-to-icon-rail (the ⇤ toggle is drawn but static in the mock), and what "Recent conversations" persists (last N brains talked to). The mock shows the evolved state, not the collapsed one.
  • Matched-results ranking is illustrative — the "why matched" lines + best-match ordering are placeholder; the real matching/ranking logic is Ben's. Categories mirror staging's chip row.
  • Want me to spec any single flow deeper (e.g. the collapsed/mobile sidebar states, or the full add-to-app states: pre/consent/connected/error)? Say the word.

Marnji · Iter 2.1 — boxless lane + pre-answers swipe stack + real bookings (Khoa's notes) · design-only, for build by Ben · Sana · 2026-07-24 · iterates on the R2 set, grounded in the current staging.marngi.com build (/chats, /pre-answers, session model).