Holy Trinity Caisse — Public Dashboard Specification

"For a caisse that hides its ledger has already ceased to be one." — Charter, §The Promise

Purpose

A single public page — no login, no analytics, no ads — that renders the numbers a skeptic can use to verify the caisse is real, solvent, and doing what it claims. Refreshed nightly from the ledger. Anyone can bookmark it and check.

URL (proposed): caisse.holytrinity.example.org/ledger (or holytrinitycaisse.example.org for a standalone shell)

The Seven Proof-Metrics

Each metric has a headline number, a plain-English definition, a target band, and a sparkline of the last 12 weeks.

1. Token Velocity

Median days from token earned to token spent.

• Proves: Deeds are being done and redeemed. Tokens are not hoarded (theatre) or dying in dormant accounts (shop doesn't stock what people need).
• Target band: 30–90 days.
• Alert if: > 180 days (catalog problem) or < 7 days (possible round-tripping — investigate).
• Displayed as: "Median token lifespan: 47 days. Half of tokens earned are spent within 47 days of the deed."

2. Fund Solvency Ratio

Dollars in the pooled fund ÷ outstanding token liability.

• Proves: The caisse can settle every token in circulation with a vendor payment. Below 1.0, the caisse is technically insolvent.
• Target band: ≥ $1.00 per outstanding token.
• Alert if: < $0.75 (call the stewards) or > $3.00 (donors outpacing deed-earners — expand catalog or witness network).
• Displayed as: "Solvency: $1.24 per token. For every token in a parishioner's account, the fund holds $1.24 to settle it."

3. Procurement Spread

(Retail dollar value delivered − Actual dollars paid to vendors) ÷ Retail dollar value delivered.

• Proves: The Amazon Business + ATEP + OMNIA + rebate stack is delivering the claimed savings. This is the parish's version of a caisse populaire's interest spread.
• Target band: ≥ 15%.
• Alert if: < 10% (renegotiate with account executive) or > 30% (verify math — likely category-mix issue).
• Displayed as: "Spread: 17.3%. For every $100 of goods delivered at retail value, the caisse pays $82.70 — the difference is stewardship, not profit."

4. Deed Diversity Index

Count of distinct earner-parishioners in the trailing 90 days.

• Proves: The caisse is a network, not a household. Diversity is what makes it catholic (small-c: universal).
• Target band: ≥ 20 in the first 6 months; ≥ 50 by end of year 1.
• Alert if: < 10 for two consecutive months (witness network too small).
• Displayed as: "Contributors: 34 parishioners. Distinct people who earned at least one token in the last 90 days."

5. Household Participation Rate

% of parish households that routed at least one certified purchase through the caisse in the trailing 90 days.

• Proves: The parish has actually shifted its consumer behavior. This is the metric that says lane 1 (certified shopping) is real, not aspirational.
• Target band: ≥ 40% by end of year 1; ≥ 60% by end of year 2.
• Alert if: < 20% at end of quarter 2 (the certification funnel is broken; steward outreach needed).
• Displayed as: "Participation: 43% of households. 187 of 434 registered households shopped through the caisse in the last 90 days."

6. Deed Video Coverage

% of deeds in the ledger that carry a linked video record (unlisted or public) on at least one witness platform.

• Proves: The caisse doesn't just count deeds; it can show them. This is the honesty metric — the check against a ledger inflated by paper attestations.
• Target band: ≥ 85% by end of year 1.
• Alert if: < 60% (evidence workflow is failing; retrain witnesses).
• Displayed as: "Video-witnessed: 91% of deeds. 374 of the 412 deeds this quarter carry a video record."

7. Public Witness Revenue

Dollars flowing to the caisse fund from public witness platforms in the trailing quarter, itemized by platform.

• Proves: Lane 3 is real. The parish's public witness is not merely propagation; it is a self-funding surface.
• Target band: Any positive figure in year 1; ≥ $10,000/year by end of year 2 (single-platform threshold).
• Alert if: No revenue for three consecutive quarters after channels reach eligibility (propagation cadence is broken).
• Displayed as: "Public witness revenue this quarter: $1,687" — with per-platform breakdown table:

The Page, Written Out

────────────────────────────────────────────────
  HOLY TRINITY CAISSE — Public Ledger
  Updated nightly · Last refresh: 2026-07-18 23:00 CDT
────────────────────────────────────────────────

  Token lifespan          Solvency
  47 days                 $1.24 per token
  ▁▂▃▃▄▅▄▅▆▅▆▇            ▄▅▅▆▆▇▇▇▆▇▇▇

  Procurement spread      Contributors (90d)
  17.3%                   34 parishioners
  ▂▃▄▄▅▆▆▇▇▇▇▇            ▁▂▂▃▃▄▄▅▅▆▆▇

  Household participation Video coverage
  43% of households       91% of deeds
  ▂▃▃▄▄▅▅▆▆▆▇▇            ▄▅▅▆▆▇▇▇▇▇▇▇

  Public witness revenue (this quarter)
  $1,687   [YouTube $847 · X $312 · Meta $528]

────────────────────────────────────────────────
  Tokens earned this quarter:     412
  Tokens spent this quarter:      378
  Certified households:           434
  Certified purchases (Q):        1,247 · $18,340 retail
  Spread returned to fund (Q):    $3,162
  Dollars donated (Q):            $9,840
  Public witness revenue (Q):     $1,687
  ─────────────────────────────
  Total fund inflow (Q):          $14,689
  Dollars paid to vendors:        $12,203
  Retail value delivered:         $19,285
────────────────────────────────────────────────

  Full charter · Quarterly report (PDF) · About PAIX
────────────────────────────────────────────────

That is the whole page. No dashboards behind logins. No aggregate charts of individual behavior. Parishioner names are never on this page — only counts.

The Public Deed Wall

A separate page at caisse.holytrinity.example.org/deeds renders the wall of publicly-consented deed clips as a grid of thumbnails, each 20–90 seconds, each with:

• Deed category (elder transport, tutoring, meal delivery, etc.)
• Month and year
• The witness's name (never the beneficiary's, unless they explicitly opted in)
• A click-through to the clip on its home platform (YouTube for the durable archive; X and Meta as propagation surfaces linked from the same source)

Only deeds with youtube_privacy = 'public' AND beneficiary_consent_given = 1 appear on this wall. Unlisted deeds still count in ledger aggregates but do not appear.

The wall is the evangelism surface — the place a Grand Knight sends his brothers on Monday morning, the place a bishop's communications office finds material, the place a mother in Missouri sees her son's Knights of Columbus witness clip.

Data Sources

• Ledger DB (SQLite, holy-trinity-caisse repo) — deeds, witnesses, redemptions, transfers, certifications.
• Vendor invoice log (Amazon Business exports + local merchant receipts) — actual dollars paid.
• Donor CRM (Stripe + manual entries) — dollars in.
• Catalog (shop.config.json) — retail prices, token prices, actual expected prices.
• Platform revenue exports (YouTube AdSense CSV, X Creator monetization report, Meta Creator Studio) — imported monthly into revenue/YYYY-MM.json.

A nightly Node script recomputes all seven metrics and writes them to public/ledger.json, which the static page reads.

What is not on the dashboard

Deliberately absent:

• Individual parishioner balances or names. Ever. The witness ledger is stewarded, not published.
• Beneficiary faces in aggregate imagery. The Public Deed Wall shows only doer-and-witness clips or auto-blurred beneficiary clips.
• Donor amounts by name. Aggregate only, unless the donor explicitly consents to be listed.
• Growth vanity metrics. No "signups," no "engagement," no MAU. The caisse is not a platform.
• Charity-porn imagery. No "meet the family we helped." Dignity is structural, not decorative.

What this dashboard enables

When a Grand Knight, a bishop, an auditor, or a skeptical journalist visits caisse.holytrinity.example.org/ledger, they can — in under thirty seconds — determine:

1. Is anything actually happening? (Token velocity, contributors, participation.)
2. Is it solvent? (Fund solvency ratio.)
3. Is the economic thesis real? (Procurement spread, spread returned to fund.)
4. Is it a real network, or a two-family favor economy? (Deed diversity, participation rate.)
5. Can the deeds be seen, or only claimed? (Video coverage.)
6. Does the public witness pay for itself? (PAIX Witness revenue.)
7. Is the parish's professional class contributing, and at what personal cost? (Professional services delivered + average walk-away.)

If those nine numbers hold for two quarters, PAIX has proven its economic model. That is the entire point.