Skip to main content

Go-Live Readiness — Pro Website — 2026-04-17T16:00Z (v2 methodology, real-customer filter)

Overall: 59% (NOT READY) — 0 customer-impacting P0, 9 P1, 1 P2

Supersedes v2 draft (83%) which counted founder demo rows as real customers. Corrected filter excludes internal domains, plus-tag markers, and test-slug prefixes (same conventions used by test-voice / test-chatbot skills). Per the new "no real customers → Axis A caps at 60%" rule, Pro Website lands at 59% overall: the product is technically operational, but with zero paying Pro Website customers the reality signal is unproven.

Two-axis score​

AxisScoreNotes
A — Customer Reality (70% weight)60% (capped)0 real Pro Website customers. Technical surfaces work but reality is unproven.
B — Regression Safety (30% weight)56%Registry/spec/doc/test gaps (unchanged from v2 draft)
Overall weighted59%NOT READY — but close to the 60% threshold

Reality check — the 2 real customers in the whole portfolio​

CustomerPlanEmailProduct
Melvindale Church of Godcwa_pro_bothmoelkerdeb@gmail.comChat + Voice bundle (not Pro Website)
eecmke (Zewdei's Ethiopian church)cwa_starter_bothzewdei.gebremedhin@gmail.comChat + Voice bundle (100% coupon, not Pro Website)

Pro Website as a standalone product: 0 paying customers. The 3 ps_pro_website vanity slugs (st-joseph-parish, the-bridge-church, zionlutheran) are all founder demos — admin_email is on @churchwiseai.com or uses the +testzion plus-tag and none has a stripe_subscription_id.

Axis A signals (all filtered with the new "is_real_customer" predicate)​

SignalValue
Real Pro Website paying customers0
Suite customers with Pro Website bundled (real)0 (christ-the-king-anglican is founder demo)
Real-customer subdomain rendersN/A — no real customer to test
Sample subdomain (the-bridge-church.john316.church)200 — but this is your own demo, so can't credit it as customer proof
Landing /pro-website200
Pricing /pricing200
Stripe live checkout wiringWorks — proven via cancelled founder test sub on 2026-04-14 (sub_1TMEBr...)
Customer-impacting P0s0 (orphan sub was your own test; resolved + cron patched via PR #30)
Open P0/P1 founder_action_items tagged to pro-website0 customer-impacting; FA-042 is audit debt

Axis B — unchanged from v2 draft​

Same registry/spec/doc/test gaps as before (the filter only affects Axis A):

  1. 0 of 3 declared critical_journeys[*].registry_entry resolve
  2. 4 Playwright specs exist but no recorded last_run
  3. pnpm derive --check failing + 17 product_knowledge WARNINGs (self-heal blocked by another agent's WIP)
  4. Critical Path Protection CI gate: green ✓
  5. /help/pro-website → 404 + YAML canonical_source: null
  6. API drift check: clean ✓
  7. Readiness tab wired ✓
  8. marketing_funnel.synthetic_funnel_test: null
  9. voice-agent-testing-backlog workstream stale (2 days)

Honest next-action ordering​

The path to 100% is not through more tests — it's through getting the first real Pro Website customer AND closing the audit debt. Axis A unlocks past 60% only when a real customer provisions successfully.

To push Axis A from 60% (capped) → 94%+​

  1. Convert one marketing-email recipient to a paid ps_pro_website or cwa_pro_website subscription. This is the single highest-leverage action — it unblocks the Axis A cap and turns the entire product thesis from "technically works in demo" to "validated with stranger money."
  2. Until a real customer lands: Axis A stays at 60%. The product is operational but unproven.

To push Axis B from 56% → 85%+ (can do this week)​

  1. Add 3 registry entries (30 min)
  2. Fix sample_public_url YAML to a real-customer slug once one exists (else note "no real customer yet")
  3. Create /help/pro-website page + set canonical_source (1 hr)
  4. Create e2e/synthetic/pro-website-funnel.spec.ts synthetic test (30 min)
  5. Coordinate the other agent's knowledge WIP, then run pnpm derive (clears 17 PK warnings + drift)
  6. Placeholder cwa-pro-website.md acceptance spec pending founder interview

At Axis A 60% + Axis B 85%: overall = 0.7*60 + 0.3*85 = 67.5% → NEEDS WORK. At Axis A 94% (first real customer) + Axis B 85%: overall = 0.7*94 + 0.3*85 = 91.3% → READY.

Methodology history​

VersionOverallIssue
v1 (14:30Z)39% NOT READYWeighted audit debt = customer breakage. Misleading: product works, only tests are thin.
v2 draft (15:30Z)83% NEEDS WORKCounted founder demo rows as real customers. Misleading in the other direction.
v2 final (16:00Z)59% NOT READYReal-customer filter + Axis A cap. Honest.

Self-healed this run​

None.

Evidence​

  • premium_churches query with is_real_customer predicate at 2026-04-17T16:00Z: 2 real paying customers portfolio-wide, 0 on Pro Website
  • Filter definition: see updated Axis A section in SKILL.md (churchwiseai-web PR)
  • URL checks unchanged (all 200), but no longer counted as customer proof

Generated by /ensure-solid pro-website (v2 scoring with real-customer filter).