Open questions for Lejla
Sixteen questions. Fourteen came from building the flow map; two more (Q15–Q16) were opened by our July 27 app review, which also answered seven of the originals — Q1, Q2, Q3, Q4, Q5, Q8, and Q9. Five of those overturned the default we had proposed (they were simply wrong); Q3 was confirmed and Q9 refined. The answered ones are marked Answered July 27 below, with what changed. The rest still carry a proposed default (correct if wrong) or wait on your confirmation.
The August 5–6 flow-map review with Lejla and Yasmine answered one more — Q13, recurring weekly booking, which is now decided in favour of building it. What remains open there is the booking policy behind it, not the screen.
The August 11 2026 round — the coach-app session of August 6, Lejla's coach-side audit, and the review of the v12 change document — closed most of what it touched and parked eighteen decisions that are not ours to take. They are collected at the foot of this page, the three blocking ones first, each carrying the number it holds in the review register.
New students & placement
It's a recorded submission, not a booked 1:1. The student moves through a short series of graded reading screens — letter recognition → words → a short surah everyone knows → a more technical ayah — recording themselves. The recording routes like a normal submission to a pooled first-assessment TA account (TAs share credentials and rotate so everyone is assessed quickly), and they place the level from it. Unchanged in the revamp.
Changed: overturns the old “booked 1:1” default — the assessment reuses the weekly submission flow.
Still to verify: the exact reading stages against the real first-assessment screens (screenshots or dummy credentials coming from you); the student-facing rename away from “assessment” is #16 below.
Placement is student-paced — days to about three weeks (people often rehearse for a couple of weeks before submitting), with no 48-hour promise. And waiting isn't a quiet hold: a not-yet-placed student is Level 0, a real status whose interface is the full app minus lessons — the pending recitation shows as a submission awaiting feedback, and they get assigned Level-0 live classes and resources. Students who never submit get reminder emails, then are auto-placed into Level 2.
Changed: overturns both the 48-hour SLA and the “quiet waiting screen” default.
No. New students place into a Year-1 level only; the only route into Year 2 is graduating Level 4.
Confirmed as proposed.
A welcome email with credentials — email + password (the system that's worked for ten years). Sign-in is email + password, not a magic link. The sign-in screen doesn't mention enrollment, and the “Enrollment happens at almaghrib.org” footer comes out — App Store rules restrict pointing to an external sales page.
Changed: overturns the magic-link-primary default.
Year 2
Plan-based, bought at checkout — not universal. The default is 1 per week for every elective; only Tajweed Practice (the lessons-free elective) sells 2–3 per week today. The UI rule: communicate what the student signed up for (“Your plan includes N appointments a week”), with more arranged through admin.
Changed: overturns the “up to 3, included in tuition” default. Opening a 1/2/3 choice to all electives at checkout was floated and not adopted.
No — removed entirely in v1. Today — Year 2 has no record button. We'd only revisit if TAs want short pre-appointment recitation clips.
The data model already defines this: cancellation is guilt-free up to the last minute (frees the slot, no penalty); a no-show is joining under 75% of the session or not at all; two consecutive no-shows auto-escalate by email to program staff. No in-app progress penalty.
Just confirm this carries unchanged to Year-2 multi-TA bookings.
No — Year 2 lessons are per-elective. Every Year-2 course has its own syllabus and lessons (that's the point of electives); Tajweed Practice alone has no lessons (appointments only). The only universal Year-1→Year-2 difference is the feedback format — appointment notes instead of submission feedback.
Changed: overturns the “same lessons for everyone” reading; the orientation screen's “weekly lessons are the same as ever” line was wrong.
Yes — build it. Lejla, on the flow map: “The previous version of this map showed the option to book the same appointment every week for the next 15 weeks. I love that… we need an additional button that allows a person to book every Thursday at 6pm up until the end of the semester.” Yasmine agreed. It behaves as a bulk action that produces individually cancellable appointments, not a locked string — each week can still be swapped or cancelled on its own.
Changed: overturns our proposed default. We had dropped the recurring option as redundant and marked the screen for retirement; it is back, and the flow map now carries it as decided rather than open.
What is still open is policy, not the screen, and it is addressed to us: whether students can hoard later slots, whether a slot is released after two no-shows, whether the current 24-hour no-rebooking rule becomes a 48-hour cancellation notice, and whether recurring booking is offered in week 1 only. The live copy (“even a minute before frees the slot”) does not match the real rule today and needs correcting either way.
Added July 27, from the ambient-layer pass. Our default: outside the app, appointments appear as facts — the next one, named and timed (“1:1 · Ustadh Kareem · Thu 6:00 PM”), plus a quiet “Book another” action while the student's plan still allows one that week. We would never show a remaining count (“1 left this week”) — on a Lock Screen that reads as a nudge to use up the allowance, which cuts against the no-manufactured-urgency rule.
Tell us if you'd rather widgets not mention booking at all — or, the other way, if you want the remaining count surfaced.
Returning students
Auto-enrollment applies to level graduates: a level graduate rolls into the next level with auto-billing, and a Level-4 graduate auto-enrolls into Year 2's mandatory first semester (an email, no checkout). Year-2 elective sign-ups and repeat students go through web checkout, not auto-enrollment.
Refined: “returning students are auto-placed” is really “level graduates auto-enroll; electives and repeats check out.” Edge case unchanged — repeating Year 1 / re-assessment is handled off-app in v1.
Today shows a semester-complete state; re-enrollment happens on the web like all sales; the app flips the account's status once the new semester is purchased.
This is the default we're least sure of — worth a real answer.
Coach / TA
One pool. The role-gated coach section shows both the review queue and the schedule; a coach who only does one kind simply has an empty other queue. The shared student record (visible to all coaches) already supports rotation.
The designed flow is record → transcribe → auto-summary → the TA edits → send. The TA's edit is the approval — no second approver.
New to confirm — opened by the July 27 answers
Two specifics the July 27 corrections themselves raised — smaller than the questions above, but each one shapes copy or a screen.
We're reading the transcript's garbled “advanced weed rules / advantage read walls” as “Advanced Tajweed Rules” — the mandatory first Year-2 semester a Level-4 graduate auto-enrolls into. This is distinct from the appointments-only “Tajweed Practice” elective — please don't let us conflate the two. Confirm the exact name before it appears in any UI; until then it stays flagged “pending confirmation” wherever it shows.
Since people “don't like the word assessment and just don't submit it,” our proposal — already encoded in the mockup — is the student-facing name “Your first recitation”; the internal / coach side keeps “First assessment.” Confirm or correct the wording.
Folded in: the auto-placed placement-result tone. For a never-submitter auto-placed into Level 2, the result screen shouldn't celebrate a grading that never happened — proposed: “We've placed you in Level II to get you started; your coach fine-tunes from your first weekly recitation.” Confirm the tone.
Parked by the August 11 2026 round
Eighteen decisions the August 11 adjudication deliberately did not take, because they belong to the program rather than to the design. The first three block build work — two of them are drawn on the flow map as frames marked PROPOSAL — DECISION NEEDED, and nothing else depends on them. The rest can wait, but each one will otherwise be decided by accident. The number in the margin is the item's number in the review register (reviews/v13-review-response/unified-register.json), so any answer can be traced back to what was actually said.
Five of the eight coaches surveyed asked for a way for a student to reply to feedback. Against it: the objection, made in the room, that a reply channel turns feedback into a conversation that then has to be graded — and the workload it adds is unscoped (U-306, below). It is drawn as C12 on the flow map: a short reply, not graded, not re-scored.
Either answer is cheap today and expensive in three months. If it is in, it needs a coach-workload answer with it.
Eight of eleven coaches teach live and have no surface for it at all — no roster, no attendance, no way to notice the student who never speaks. It is drawn as C13: roster, attendance, a timer and a speaking-turn tracker. Whether that is a v1 feature or a later one is a program call, not a design one.
The direction is decided: coaches log hours in the app (C16), all three work types, correctable with a one-line reason until the period closes. What is not decided is what happens after that — whether finance takes the export and Time Clock Wizard is retired, or the two run in parallel. The all-or-nothing warning stands: a half-migrated hours system is worse than either whole one.
Year 1 messages the coach she was given; Year 2 messages the teachers she has booked. Which tab that lives under — and whether the coach's inbox (C15) and the student's messages (S45) are one surface or two — is open. For v13 both stay where they are.
Every screen that says “due Sunday” assumes she is on the current week. A student two weeks behind needs either a moving date or an honest one; the app cannot invent the rule.
S15 asks when a student can be somewhere live. Whether that means a day, a half-day, or an hour band changes both the screen and what scheduling can do with it.
The copy on the booking screens is wrong today in both directions. The two-no-show rule is encoded with a notification and an appeal path; the cancellation window itself still needs a number.
C04 now carries a pass/fail control that stays padlocked until the end-of-course evaluation opens. What makes a pass is a program answer, and it is the thing the control is waiting for.
The EOC is referenced everywhere and defined nowhere in the app. If there is a sheet, it is a screen; if there is not, the padlock above opens onto a judgement call.
C11 gives a TA four stages and a level control. Whether she is applying a written rubric or her judgement changes what that screen should show her while she listens.
Covered, work on, what is next are decided and drawn on C08. If anything else is mandatory — a duration, a rating, an attendance reason — it belongs in that form and not in a coach's memory.
C04 was rebuilt as a workload view on the evidence available — overdue, due this week, days since a student was last heard from. The deeper redesign waits on coaches, not on us; the audit was explicit that this is the one screen it could not answer without them.
C08 shows a transcript affordance inherited from the earlier design. Whether sessions can be transcribed at all — cost, accuracy on recitation, consent — is unanswered, and the note model does not depend on it.
Raised in the Aug 11 meeting as a platform question. It does not change a screen, but it changes what the lesson player can promise about offline and about playback speed.
It replaced the enrollment footer on S01 for App Store reasons. Whether it opens a support form, an email, or a help article is undecided — and it is the first thing a locked-out student taps.
S26 shows the teacher's spoken note with anchored points beneath it. Whether those points are written by her or drafted for her decides whether the screen needs to say so.
Reported by coaches in the Aug 6 session, about the app that exists today. It is an engineering diagnosis on the live product, not a redesign decision — but it is the reason some coaches distrust in-app tools at all, so it shapes adoption of everything above.
The half of U-295 nobody has scoped. A reply is only cheap if it is bounded — one per feedback, no thread, no expectation of a reply to the reply. If the answer to U-295 is yes, this needs a number before it is built.