Implementation inventory

• Four-step mobile-friendly intake at /7-figure-readiness: profile, dated business snapshot, 27-item owner self-assessment, confirmation.
• Owner scorecard with stable keys (offer_01_owner … ceo_mindset_03_owner); blank = unknown, 0 valid.
• Advisor dashboard at /advisor: separate advisor scores + private evidence per item, verified-items counter (out of 27).
• Scoring engine: 9 area averages, overall readiness %, top-3 gaps ranked lowest-first with tie-break in canonical order, "Optimization priorities" label when all ≥ 3.
• Revenue model: monthly pace, gap, wins, held, bookings, qualified leads, paid spend, capacity-shortfall flag, "Needs input" guardrails.
• 90-day plan: three editable sprints, advisor first-sprint confirm/override with reason, sales-path selection, plan approval.
• Client-facing report at /report/$assessmentId — excludes private evidence and advisor-only notes.
• Authenticated calculation endpoint at /api/readiness/calculate (the external calculation service contract).
• Local demo store preserves dated assessments per contact so returning clients keep earlier baselines.

• Pipeline "7-Figure Readiness" with stages: New Intake → Advisor Review → Plan Approved → Kickoff Booked → Active Program → Completed.
• Custom fields on the contact/opportunity (see field list below).
• Intake form mapped from the four steps to the custom fields, posting to the create-contact + create-opportunity action.
• Workflows (intake submitted, advisor review completed, plan approved, kickoff booked) — kept in draft/disabled until approved.
• Calculation webhook connection (Make / n8n / server-side) calling /api/readiness/calculate.
• Secret VULPINE_GAZE_CALC_SECRET stored server-side via the secrets tool; the caller sends it as Authorization: Bearer ….

Native workflows cannot reliably validate/rank 9 scores or run the revenue model, so an authenticated Make/n8n/server endpoint calls:

POST /api/readiness/calculate
Authorization: Bearer <VULPINE_GAZE_CALC_SECRET>
Content-Type: application/json

{
  "assessmentId": "VGA-...",
  "scorecard": {
    "ownerScores":     { "offer_01_owner": 3, ... },
    "advisorScores":   { "offer_01_advisor": 3, ... },
    "advisorEvidence": { "offer_01_evidence": "Reviewed 3 paid pilots...", ... }
  },
  "baseline":  { "monthlyRecognizedRevenue": 60000, "avgNetRevenuePerWin": 8000, ... },
  "scenario":   { "annualRecognizedRevenueTarget": 1000000,
                  "closeRate": 0.25, "showRate": 0.75,
                  "qualifiedLeadToBookingRate": 0.5,
                  "maxSustainableNewStartsPerMonth": 7 }
}

Response (200, or 422 when review incomplete / input missing):

{
  "assessmentId": "VGA-...",
  "ok": true,
  "scoring": {
    "overallReadiness": 58.3,
    "areas": [ { "area": "sales", "score": 0, "rounded": 0 }, ... ],
    "topThreeGaps": [ { "rank": 1, "area": "sales", "firstAction": "...", "suggestedKpi": "...", "suggestedRole": "Sales Lead" }, ... ],
    "isOptimization": false, "verifiedItems": 27, "totalItems": 27, "valid": true, "issues": []
  },
  "revenueOutputs": { "monthlyRevenuePace": 83333, "newWinsRequired": 8,
                      "heldAppointments": 32, "bookings": 43, "qualifiedLeads": 86,
                      "capacityShortfall": true },
  "funnelFlags": []
}

Idempotency: dedupe on assessmentId + approvalVersion. The response header X-Dedupe-Key carries{assessmentId}::calc.

1. Intake submitted. Trigger: form submission. Actions: create/match contact → create one opportunity (dedupe on assessment_id) → assign advisor → create one review task (attach funnel flags) → send intake confirmation only via an eligible, consented channel.
2. Advisor review completed. Trigger: advisor marks review done. Actions: call calculation webhook → validate 27 scores+evidence → write back 9 area scores, readiness %, top-3 gaps, revenue outputs → keep opportunity in Advisor Review until plan approved. On failure: log assessment_id, surface error to advisor, block the results message.
3. Plan approved. Trigger: approval_version increments. Actions: move to Plan Approved → send one client summary + kickoff booking link via eligible channels (dated baseline, top-3, first sprint, KPI target; exclude private evidence). If scores/evidence change after approval, set requires_reapproval before resending.
4. Kickoff booked. Trigger: kickoff booking confirmed. Actions: move to Kickoff Booked → assign internal kickoff checklist → schedule internal review tasks at days 30/60/90 from the confirmed start date.

Every action dedupes on assessmentId + action/approvalVersion. Respect channel consent, opt-outs, and DND. Failed calculations log the assessment ID and block the results message until resolved.

• Scores of 0 count as valid; blank/fractional/out-of-range do not.
• Sales 0, Messaging 1, Visibility 2, others 3–4 → returns Sales, Messaging, Visibility in that order.
• Ties follow canonical category order.
• $1M target, $20K MRR, $8K/win, 25% close, 75% show, 50% booking → 8 wins, 32 held, 43 bookings, 86 qualified leads; capacity 7 → shortfall.
• Replayed submission/approval sends no duplicate message and creates no duplicate task.
• Opted-out contact receives no message through the opted-out channel.
• Returning client can start a new assessment; earlier baseline stays accessible.
• Client-facing report contains no private advisor evidence.

1. Dummy coaching business — open /7-figure-readiness: owner "Pat Coach", email pat@coachdemo.test, business "Apex Leadership Coaching", type Coach, baseline last month. Enter $60K recognized revenue, $8K net/win, 4 active clients, 200 unique leads, 100 qualified, 50 booked, 38 held, 9 wins, $5K MRR, capacity 7. Self-score all 27.
2. Open /advisor, open the record. Leave 2 evidence notes blank → confirm it stays in Advisor Review and verified count < 27.
3. Fill all 27 advisor scores: set Sales 0, Messaging 1, Visibility 2, all others 3–4, with nonblank evidence → verify top-3 = Sales, Messaging, Visibility.
4. Revenue model: target $1M, planned MRR $20K, net/win $8K, close 25%, show 75%, booking 50%, capacity 7 → verify 8 wins / 32 held / 43 bookings / 86 leads + capacity shortfall.
5. Confirm first sprint priority, pick "Coach · premium offer requiring diagnosis", fill the three sprints → approve plan → verify stage moves to Plan Approved and approval_version increments.
6. Open /report/$assessmentId → verify it shows baseline, top-3, first sprint, KPI, and NO private evidence.
7. Dummy local service business — repeat as "Sam's Plumbing", type Local Service, baseline last month, with a referral funnel. Pick the "Local Service" sales path. Verify the report renders the local-service steps.
8. Re-submit the same contact → verify a new assessment_id is created and the earlier baseline is still listed.
9. Toggle a communication preference off and replay an approval → verify no duplicate message/task is created.
10. Only after all pass: enable the four workflows one at a time, starting with Intake submitted.

• Intake submitted (outbound confirmation)
• Advisor review completed (writeback + results message)
• Plan approved (client summary + kickoff link)
• Kickoff booked (checklist + day 30/60/90 review tasks)

These remain draft/disabled until you approve them in the sub-account. The app-side calculation, scoring, plan, and report all work now in the preview.