* feat(billing): enforce free-plan student cap across the app
Free (BASIC) plan orgs were able to exceed the advertised 20-student
limit because the cap existed only as dashboard cosmetic copy with no
server-side enforcement.
Add a single source of truth (getStudentLimit / STUDENT_LIMITS) and a
central guard (assertStudentCapacityOrThrow) wired into every path that
creates a new student-role org member: self-enrolment, course-invite
acceptance, bulk audience import, and org-invite acceptance. The guard
throws UPGRADE_REQUIRED (403) and notifies org admins by email (throttled
to once per org per day) when the limit is hit.
Surface the limit in the UI: disable the audience import button and the
course new-student invite section with upgrade banners for admins, and
show a "contact your administrator" message on the public enrol page for
students (who cannot upgrade). Expose a studentLimitReached flag from the
public course endpoint so the enrol button can be disabled proactively.
Add a one-time (manual, non-cron) db script,
db:notify-students-over-limit, to email admins of orgs already over the
limit at rollout.
* feat(billing): guard remaining student-creating paths against the cap
The initial cap enforcement missed several service entry points that also
create a new student-role org member. Add the capacity guard to:
- autoEnrollStudent (public org self-join on a tenant site)
- cohort enrollment (bulk group enrollment and single member accept)
- ensureProgramCourseAccess (program/cohort-granted course access)
For the cohort and program paths, only genuinely new members are counted
(existing org members are skipped) so re-enrolling current students is
never falsely blocked. The core path enforces via the shared limit check;
it throws UPGRADE_REQUIRED without the admin email since the queue lives
in the api layer.
* fix(billing): don't show cap block to existing org members on enroll page
The public course-by-slug endpoint is unauthenticated, so the enroll
page's studentLimitReached flag is org-wide and can't tell whether the
current viewer is already an org member. An existing student of a full
BASIC org would see Join disabled even though the backend correctly lets
existing members enroll without consuming a new seat.
Only pre-block anonymous visitors (would-be new joiners). Logged-in users
are governed by the membership-aware backend, which allows existing
members and returns UPGRADE_REQUIRED (surfaced via the enroll onError
mapping) for genuinely new students.
* feat(billing): deliver org plan-limits via /account with an extensible registry
Replace the ad-hoc, per-surface audience fetches (each calling getOrgAudience
just to read a count) with a single, future-proof delivery of org usage +
limits on the /account payload, gated to admin/tutor members (students never
receive limit data; a plain COUNT, no student rows).
Model it generically so future org limits (courses, etc.) drop in without
reshaping the payload or re-plumbing the frontend:
- New packages/utils plan-limits registry: PLAN_LIMIT_RESOURCES, PLAN_LIMITS
(reuses STUDENT_LIMITS), getPlanLimit, and a JSON-safe ResourceUsage /
OrgUsageLimits shape with isResourceLimitReached / resourceLimitRemaining.
- getAccountData attaches a per-org `limits` map for admin/tutor orgs.
- Store selectors currentOrgLimits / studentUsage / isStudentLimitReached;
sidebars, upgrade card, People pages, and invite modals read these instead
of fetching audience rows.
Bring cohort surfaces to parity with course: cohort People page banner and
cohort sidebar People upgrade icon, matching the course equivalents.
Also: close the upgrade modal that opened over the invite modal (drop the
`add` param via openUpgradeModalOver), redesign the sidebar upgrade card
(progress toward the student limit), add a Progress `variant`, and document
the no-complex-expressions-in-object-literals rule in AGENTS.md.
* refactor(billing): address PR review on the limits work
- Dedupe new student profile IDs before the cohort seat check so duplicate
batch entries can't overcount against the cap (Greptile P1).
- Drive the org-sidebar upgrade indicator from a dynamic nav-item `upgrade`
flag (audience config declares `upgradeResource: 'students'`; groups set it
from a `limitsReached` map) instead of a hardcoded path check.
- Collapse the audience Import button into a single button with ternary
`disabled`/`href`.
- Simplify the notify-students-over-limit upgrade link to app.classroomio.com.
- Drop the comment on the org `limits` field.
- Add .greptile/rules.md documenting the manual one-time db/src/scripts
convention (operator-controlled reruns) so idempotency isn't reflagged.
* fix(billing): name the org in the enroll capacity message
Change the enroll-page student-limit message from "This course has reached
its student capacity" to "{orgName} has reached its student capacity", since
the cap is org-wide, not course-specific. Pass orgName from the loaded org on
the page and from the currentOrg store in the enroll onError snackbar.
* style(billing): use PremiumIcon for the Authentication upgrade indicator
Replace the RocketIcon on paid settings sub-items (e.g. Authentication) with
the PremiumIcon used on the Audience item, right-aligned (ml-auto) to match.
ClassroomIO Backend
Supported Features
- Certificate download for students.
- Uploading videos to cloudflare (for storage) and Muse.ai (for transcription)
Running locally with video upload + playback
Lesson videos run through an HLS pipeline on Cloud. The browser encodes a
360/720/1080 ABR ladder with Mediabunny, uploads segments to R2 via presigned
PUTs, and the player loads /hls/{assetId}/master.m3u8 from the API.
How /hls/* resolves depends on the environment:
| URL the browser hits | Who serves the bytes | |
|---|---|---|
| Cloud production | https://<tenant-host>/hls/... |
Cloudflare Worker (apps/tenant-router) — intercepts the path and streams directly from an R2 binding. Render is never in the bytestream. |
| Local dev / self-hosted | ${PUBLIC_SERVER_URL}/hls/... (direct to API) |
Hono streaming route on the API — session-auth, then getS3Client().send(GetObjectCommand). |
Self-hosted (PUBLIC_IS_SELFHOSTED=true) skips HLS entirely and falls
back to raw-MP4 upload — no R2-specific code path. Everything below is for
Cloud-style local development.
Required env
apps/api/.env:
# Object storage — Cloudflare R2 OR an S3-compatible local backend (MinIO, etc.)
CLOUDFLARE_ACCOUNT_ID=…
CLOUDFLARE_ACCESS_KEY=…
CLOUDFLARE_SECRET_ACCESS_KEY=…
# (or OBJECT_STORAGE_ENDPOINT / OBJECT_STORAGE_ACCESS_KEY_ID / OBJECT_STORAGE_SECRET_ACCESS_KEY)
# Optional — only needed if you want extract-audio → Whisper transcripts
OPENAI_API_KEY=…
# Optional — Redis enables the BullMQ media job pipeline. Without it the
# asset row is still created, but no transcription/thumbnail jobs run.
REDIS_URL=redis://localhost:6379
# Only required when you want to exercise the production cookie-mint path
# (`POST /organization/assets/:assetId/hls/cookie`). Locally the Hono /hls/*
# route uses session auth, so most of the time you can leave this unset.
# Same value must also be set as a Cloudflare secret on tenant-router for prod.
HLS_SIGNING_SECRET=…
apps/dashboard/.env:
# Already required by the existing API client — used as the HLS base URL too.
PUBLIC_SERVER_URL=http://localhost:3000
Run
pnpm --filter @cio/api dev
pnpm --filter @cio/dashboard dev
# (optional) pnpm --filter @cio/jobs dev # for transcription + thumbnail jobs
Drop a video into a lesson; the dashboard encodes locally, you'll see
segments arrive under videos/{assetId}/hls/… in R2 and an assets row with
hls_manifest_key populated. Playback flows directly browser → API → R2 —
no wrangler dev needed locally.
Production setup (one-time)
For Cloud deployment, generate the HMAC signing key the Worker uses to verify playback cookies and set it on both sides:
node -e "console.log(require('crypto').randomBytes(48).toString('base64url'))"
# On Render (apps/api service env):
HLS_SIGNING_SECRET=<paste>
# On Cloudflare (apps/tenant-router):
pnpm --filter @cio/tenant-router exec wrangler secret put HLS_SIGNING_SECRET
# paste the same value when prompted
Debugging Fly.io
View logs from fly.io in real time