feat(billing): plan taxonomy reconciliation + Stripe sync + internal-tester allowlist (#164)

Co-authored-by: Michael Chihlas <michael@resolutionflow.com>
Co-committed-by: Michael Chihlas <michael@resolutionflow.com>

CURRENT-STATE.md

@@ -2,16 +2,30 @@
 
 > **Purpose:** Quick-reference file showing exactly where the project stands.
 > **For Claude Code:** Read this first to understand what's done and what's next.
-> **Last Updated:** May 1, 2026
+> **Last Updated:** May 7, 2026
 
 ---
 
-## Active Phase: Go-to-Market Validation (Pre-PMF)
+## Active Phase: Go-to-Market Validation (Pre-PMF) — Self-serve cutover (Phase O) in flight
+
+Self-serve signup backend (Phase 1) and frontend (Phase 2) are merged. Cutover (Phase O) is gated on manual ops: live-mode Stripe Dashboard config, Railway prod env vars, internal validation pass against prod test mode, then the public flag flip. Plan: `docs/superpowers/plans/2026-05-06-self-serve-signup-phase-2-frontend-cutover.md`.
 
 ---
 
 ## Recently shipped (post-0.1.0.0)
 
+- **2026-05-07 — PR #164 (open)** Plan taxonomy reconciliation + `INTERNAL_TESTER_EMAILS` allowlist. Marketing surface (PricingPage, Stripe products) used `Starter / Pro / Enterprise` while backend was on `free / pro / team`, leaving `plan_billing` unseeded and `BillingPlan` schema accepting a literal that violated the FK. Migration `4ce3e594cb87`: rename `team` → `enterprise` in `plan_limits`, add `starter` row (caps interpolated between free and pro: `max_trees=10`, `sessions=75`, `ai=15/mo`), defensive update of any subscriptions on the `team` slug. Code rename across schemas, `Subscription` paid-plan checks, admin endpoints, and frontend `useSubscription`. Resource visibility (`Tree.visibility='team'`, `StepLibrary.visibility='team'`) is a separate domain and intentionally untouched. New `backend/scripts/sync_stripe_plan_ids.py` — idempotent upsert of `plan_billing` rows from Stripe products by exact name match, picks active monthly recurring price, leaves annual fields NULL by design. Test-mode `plan_billing` populated for all 3 tiers in dev. Phase O Task 46 allowlist: `INTERNAL_TESTER_EMAILS` env var (comma-separated) bypasses `SELF_SERVE_ENABLED=false` for specific authenticated users — `Settings.is_self_serve_active_for(email)` centralizes the check; `/config/public` returns `self_serve_enabled=true` for allowlisted authenticated callers; `/auth/register` allows allowlisted emails to register without invite code. New `get_current_user_optional` dep for endpoints that work both anonymous and authed.
+
+- **2026-05-06 — PR #163** Seed test users marked email-verified. Fixed seeded users showing the email verification banner in dev/test, blocking flows that gate on `email_verified=True`. Squash-merged into main as `dad5e1f`.
+
+- **2026-05-06 — PR #162** Self-serve signup Phase 2 (frontend cutover). 18 commits across Tasks 27–44 of the Phase 2 plan: backend remainders + frontend billing foundation + auth surfaces (OAuth + accept-invite + verify-email) + welcome wizard + dashboard redesign (TrialPill, NextStepCard, unified checklist) + public surfaces (`/pricing`, `/contact-sales`) + beta-signup deprecation. Single alembic head `c6cbfc534fad` (no new migrations in Phase 2). Squash-merged as `f1be3ab`.
+
+- **2026-05-?? — PR #161** Self-serve signup backend (Phase 1). `plan_billing` sibling table for Stripe + catalog metadata, `sales_leads` and `stripe_events` tables, `complimentary` status with `has_pro_entitlement`, `BillingService.start_trial` wired into `/auth/register`, `/billing/checkout-session`, Stripe webhook handler with idempotency via `stripe_events`, Google + Microsoft OAuth callbacks with `oauth_identities` linking, `require_verified_email_after_grace` + `require_active_subscription` guards, bulk-create + soft-revoke invite endpoints, account-invite email-match enforcement, pilot complimentary backfill, `accounts.team_size_bucket` + `primary_psa` for wizard. Squash-merged as `f918b76`.
+
+- **2026-05-02 — PR #159** In-product User Guides rewrite to Diátaxis how-tos. Replaced 15 feature-dump guides with 43 problem-oriented how-tos grouped under 10 categories. Dropped Maintenance Flows / AI Assistant / Flow Assist Sparkles guides (UI no longer exists). Renamed Step Library → Solutions Library. Authored 14 net-new how-tos for FlowPilot-era surfaces (tasklane keyboard flow, what-we-know, resolve, escalate, record-fix-outcome, post-docs-to-ticket, share-update, pause-and-leave, build-script-from-scratch, open-suggested-flow, pin-a-flow, invite-teammate, etc.). Schema additions: `category`, optional `relatedSlugs`. Browser-verified against engineer + owner login.
+
+- **2026-05-?? — PR #160** Post-PR-159 UI cleanup — sidebar IA + account redesign. Squash-merged as `a8b22cf`.
+
 - **2026-05-01 — PR #158** Session-screen UX impeccable pass + tasklane keyboard flow. Heuristic score 24/40 → 33/40 across five sub-passes (distill, quieter, layout, typeset, polish). Removed duplicate "Suggested checks" chip strip → TaskLane is the single source of truth; added inline `Next steps · N pending` cue on the latest action-bearing AI bubble; consolidated session header to Resolve + Escalate + ⋯ kebab; centered messages column to match composer; dropped all banned decorations (side stripes, gradient surfaces, backdrop blur, accent borderTop) for a single decoration channel per surface; unified 14 text sizes into a 5-step scale. TaskLane keyboard flow: Enter submits + auto-advances, Shift+Enter newline, Esc cancel, focus jumps to Send after the last task. Banner ↔ script-panel are now linked (collapse hides both, any outcome closes both). WhatWeKnow section is collapsible with `sessionStorage` memory + auto-collapse-at-5-facts. Side fix: ParameterizationPreview no longer over-highlights short parameter values (word-boundary check). Two backlog entries logged in `.ai/TODO.md`: ConcludeSessionModal multi-select and `bg-card-hover` Tailwind drift in CommandPalette.
 - **2026-05-01 — PR #156** Suggested-fix "Awaiting verification" outcome. Engineers can now park a fix in `applied_pending` (waiting on client power-cycle, AD replication, license sync, etc.) instead of forcing a synchronous worked/didn't/partial verdict. PendingBanner with worked / didn't / update reason / dismiss; nudge "Still checking" records pending with a reason; page-level Resolve auto-patches pending → success before the resolution flow opens; page-level Escalate intercepts pending. Migration `c0f3a4b7e91d` (`pending_reason` column + status CHECK constraint).
 - **2026-04-30 — PR #155** Escalation Mode wedge. Magic-moment handoff-context screen for senior pickup, live SSE escalation arrivals, post-claim time-to-first-action metric (`GET /analytics/flowpilot/escalations`), atomic role-gated claim with conflict resolution, queue self-exclusion, chat ownership extended to claimed sessions. The wedge for the first paying-customer push.
@@ -215,17 +229,30 @@
 
 ## What's In Progress
 
-- **GTM Validation:** Shadow & Ship — founder uses product for 2 weeks, then hands logins to 5 colleagues
-- **Solutions Library spec:** Written at `docs/plans/2026-03-23-solutions-library-design.md`, implementation deferred to post-pilot
+- **Self-serve cutover (Phase O):** PR #164 (open) closes the last code blockers — taxonomy reconciliation + `INTERNAL_TESTER_EMAILS` allowlist. After merge, remaining work is purely manual ops: live-mode Stripe Dashboard config, Railway prod env vars, internal validation pass with Andrea Henry + 2-3 external Directors of Onboarding, then `SELF_SERVE_ENABLED=true` flip with frontend redeploy.
+- **Stripe live-mode setup:** Test-mode is fully wired (3 products, monthly prices for Starter/Pro, Enterprise sales-led, `plan_billing` seeded via `sync_stripe_plan_ids.py`). Live mode requires manual Dashboard config — same script handles seeding live IDs.
+- **GTM Validation:** Shadow & Ship — founder uses product for real MSP tickets daily, then hands logins to 5 colleagues.
+- **Solutions Library spec:** Written at `docs/plans/2026-03-23-solutions-library-design.md`, implementation deferred to post-pilot.
 
 ---
 
 ## What's Next (Priority Order)
 
+### Phase O Cutover (Weeks 0-1)
+
+- Merge PR #164
+- Stripe Dashboard live-mode setup (Products + Prices for Starter/Pro, no Prices on Enterprise, Customer Portal config, webhook endpoint with 5 events)
+- Railway prod env vars (`sk_live_*`, `whsec_*`, `INTERNAL_TESTER_EMAILS`, prod Google + Microsoft OAuth credentials, `OAUTH_REDIRECT_BASE`)
+- Run `sync_stripe_plan_ids.py` against prod backend; verify `plan_billing` has `sk_live_*` price IDs
+- Internal validation pass (9 scenarios from Phase O Task 46 plan)
+- Email pilots about complimentary status, flip `SELF_SERVE_ENABLED=true` (frontend redeploy required for `VITE_SELF_SERVE_ENABLED`)
+- PostHog dashboards + Sentry alert at >1/hour Stripe webhook errors
+
 ### Pilot Phase (Weeks 1-2)
 
 - Founder dogfooding: use ResolutionFlow for real MSP tickets daily
-- Collect feedback on copilot-first experience
+- 3 calls with external Directors of Onboarding to validate the documentation-builder thesis (cold pitch, no friendly contacts)
+- Collect feedback on copilot-first experience and self-serve onboarding flow
 - Fix issues discovered during real usage
 
 ### Post-Pilot (Weeks 3-4)