docs(plan): web signup, login & the signed-in app v1.0 (S-series)

[oratis]

What

Design doc for the next era: meetlisa.ai signup/login + the signed-in cloud app, answering four requirements — Google/Apple/email+OTP login, automatic post-signup birth, the signed-in app with the full web feature set, and GCP (Cloud Run + GCS) production deployment.

Highlights

• Dual-path invariant (owner directive): self-hosted local (no account, flagship, forever) and optional cloud — the user keeps the power. Every design below must keep both paths visible and migratable.
• Gap analysis against the already-landed B-series (accounts/billing, PRs fix(review): App Store 2.1 remediation + accounts/billing plan (B0) #259–feat(billing): B7 — rate limits, daily cap, kill switch, reviewer-account seed #267): Apple + email/password + lazy per-uid birth + Cloud Run/GCS already exist; the real gaps are Google sign-in, email OTP (+ the missing password reset), birth transactionality/ceremony, per-uid autonomy scheduling, the website entry point, and productionization.
• Seven For/Against debates with verdicts (brand vs cloud account, Google reversal, custom zero-dep auth vs hosted IdP, OTP vs link, "full features" honest definition, ceremony vs silent birth, single-instance vs Firestore).
• Milestones S1–S7; S6 (Firestore + prod project + Secret Manager) is the hard gate before public signups.

Implementation PRs follow, one per milestone.

🤖 Generated with Claude Code

[oratis]

Review — merge-ready ✅ (design doc)

Documentation only (docs/PLAN_WEB_SIGNUP_v1.0.md) — the S-series plan for web signup / login / signed-in app. No code, no risk. It's the north star for #297/#299/#301/#302/#303/#304; the adversarial reviews of those PRs are where the substance is. Good to merge whenever; suggest landing it first so the series has its reference doc on main.