CaNaMED Platform

Status: alpha

This platform is in alpha. It runs real workshops and the reference deployment is live, but it is actively developed, and things change. Concretely, if you are considering adopting it:

• Interfaces are not stable. The session/content model is still moving — the scenario format was recently replaced by independently-picked sections, and authored-content shapes may change again without a migration path.
• Expect defects in newer surfaces. The facilitator authoring tool and the section model are the least settled parts. Recent fixes addressed authored content being silently replaced by built-in defaults — a failure mode that produces no error, so verify a scenario end to end before teaching with it.
• You are the data controller. If you run your own deployment for real participants, the privacy notice, consent flow, retention jobs and ethics approval are yours to review and operate — see docs/Third_session/PBL_platform/legal/ and the retention workflows in .github/workflows/. Some operational items (App Check enforcement, API-key restrictions) are console settings this repo cannot carry for you.
• No stability or support guarantee beyond the MIT licence. Pin a commit rather than tracking main if you need reproducibility for a cohort.

Issues and PRs are welcome — alpha is exactly when feedback is worth most.

Open-source single-page web platform for running cross-cultural, problem-based-learning (PBL) workshops in clinical communication. Pairs students from two (or more) medical schools into shared online rooms, walks them through a clinical case + a roleplay + structured team decisions, and records anonymisable contributions for downstream education-research analysis.

Originally built for the Franco-Japanese CaNaMED project (Université de Caen Normandie × Nagoya University); released as a standalone platform so any medical-school partnership can adopt or fork it.

Reference live deployment: https://canamed-69785.web.app

---

Fork this for your own partnership

The engine (script.js, lib.js) is partnership- and case-agnostic; the content (case-content.js, scenario JSON), branding, and language packs are the parts you swap.

To adapt the platform for a different partnership:

1. Fork the repo on GitHub.
2. Edit docs/Third_session/PBL_platform/platform-config.js — partnership name, university list, default language, branding colours.
3. Add your scenario(s) via the in-app scenario author tool (docs/Third_session/PBL_platform/scenario-author.html) or by editing case-content.js / CANAMED_SCENARIOS directly. See CONTRIBUTING.md for the full path.
4. Add language packs if your students don't share English — edit docs/Third_session/PBL_platform/i18n.js (instructions in CONTRIBUTING).
5. Configure your own Firebase project (free Spark plan is enough for single sessions) — drop your credentials into a local firebase-config.js and deploy. Hosting + database rules are in the repo; the GitHub Actions deploy workflow needs your project's service-account secret. Step-by-step: RUNBOOK §1 — including the one-time Hosting site creation that projects created from 15 October 2026 on need before their first CI deploy.
6. Review the ethics + privacy posture in SECURITY.md and in docs/Third_session/PBL_platform/ARCHITECTURE/OPERATOR_POLICY.md — both will need to be adapted to your institutions' data protection regimes (GDPR, APPI, HIPAA, your local IRB).

Contributions welcome — bug fixes, new scenarios, new language packs, accessibility improvements. See CONTRIBUTING.md for the development workflow and CODE_OF_CONDUCT.md for community norms. Security issues: see SECURITY.md.

---

What's in this repo

Path	What it is
docs/Third_session/PBL_platform/	The single-page web platform. Vanilla HTML/CSS/JS, Firebase Realtime Database, no build step. Live at the URL above. See its own README for setup.
tests/	Node-based unit + Firebase-rules tests (run with npm test).
tests-e2e/	Playwright end-to-end suite — splash, lobby, mobile, a11y, perf, i18n, cross-tab coordination.
scripts/	Operational helpers: local dev server, synthetic-uptime probe, cost monitor, daily backup, stale-session cleanup, pseudonymisation export.
.github/workflows/	CI/CD — test, e2e matrix, firebase-deploy, synthetic-uptime, cost-monitor, backup, cleanup.

Note on the path docs/Third_session/ — this is a historical artefact from the original CaNaMED research repo where the platform was built during the project's third workshop cycle. The structure is preserved so all internal references keep working; a future PR may flatten it.

The original research data, R analysis pipeline, and curriculum materials live in a separate private repository held by the CaNaMED research team; only the platform code is open-sourced here.

---

Quick start by role

I'm a student joining a workshop

Open the operator's URL (the reference deployment is https://canamed-69785.web.app), type the session code your facilitator gave you, fill in your name + university + year + English level. That's it. Everything runs in the browser.

I'm a facilitator running a workshop

Same URL. Click "I'm a facilitator — create a session →", set a session password, pick the scenario (or paste your own), and you get a code to hand out. The admin dashboard tracks every room live; when you're done, "End session & download archive" exports the whole session as JSON.

Detailed walkthrough: docs/Third_session/PBL_platform/OPERATOR_QUICK_START.md.

I'm a developer working on the platform

The whole app is in docs/Third_session/PBL_platform/ — index.html, script.js, style.css, case-content.js, platform-config.js, firebase-config.js, database.rules.json. No npm install for the app itself, no build step. Run the tests with:

npm ci                  # exact locked tree, same as CI
npm test                # unit tests only (node --test tests/*.test.js)
npm run test:e2e        # Playwright (set up browsers first: npx playwright install)
npm run test:e2e:rules  # database.rules.json against the emulator (needs Java)

Local dev server: node scripts/serve-platform.js then open http://localhost:8080.

Push to main and the firebase-deploy GitHub Action ships hosting + best-effort database rules (requires FIREBASE_SERVICE_ACCOUNT_CANAMED_69785 secret configured for your fork; the rules step warns and continues if that account lacks Realtime Database Admin, so hosting still ships). It deploys once the E2E tests workflow concludes green on that commit — roughly 15 minutes after the merge — and refuses to ship if any required check failed.

---

Operator privacy & ethics

This platform is designed for live use with student participants who will provide identifying data (name, university, year of study, self-reported English level) and free-text answers. Every operator deploying this platform is responsible for:

• Notice + consent — the in-app lobby shows a "How your data is used" panel before any join; align it with what students are told verbally and with your institution's IRB-approved consent text.
• Privacy policy — the bundled privacy.html / privacy-fr.html / privacy-ja.html are templates with placeholders for operator-specific values (postal addresses, DPO email, retention period, complaint authority). Fill these BEFORE first live use.
• Storage — Firebase Realtime Database (Google Cloud) + per-device localStorage. Pick a region appropriate to your jurisdiction.
• Retention — there is no automatic deletion. Purge each session after the workshop (super-admin "Purge this session's data" button in the dashboard, or delete sessions/{code} from the Firebase console).
• Pseudonymisation — when exporting group answers, the admin can tick "Anonymise names in export" to pseudonymise students per room (Student A, Student B, …).
• Right to be forgotten — students who sign in with Google can delete their account from the "My account" panel.

See docs/Third_session/PBL_platform/ARCHITECTURE/OPERATOR_POLICY.md for the full operator obligations.

---

License

MIT — see LICENSE.