# MOCK-ENGAGEMENT-1 — Engagement Scope (as submitted by the buyer)

**SYNTHETIC — MOCK ENGAGEMENT.** The buyer below is invented; this document
and every artifact of this engagement carry that label. Real provider rows
carry ONLY real board/NPPES facts; the deliberately wrong values live ONLY on
synthetic rows (see `roster.json`).

| field | value |
|---|---|
| Buyer | Example Health Plan (SYNTHETIC) — a Medicare Advantage plan stand-in |
| Submitted by | "J. Compton", Directory Compliance Lead (SYNTHETIC person) |
| Date submitted | 2026-09-13 |
| Budget line | Compliance — provider directory accuracy (annual control evidence) |
| Decision it supports | The annual directory-accuracy control packet: what was checked, by whom, against which authority, and what happened next |

## The commissioned question (exact)

> For the agreed sample of 25 provider rows from our published Missouri dental
> network directory, verify each row's **provider name, licence number, NPI,
> and practice address** against the publishing authorities — the Missouri
> Board of Dentistry roster and the CMS NPPES registry. Identify every
> unsupported or conflicting assertion. Deliver signed findings we can file
> against the control, a closure receipt per finding with our owner's
> disposition, and one follow-up check on each correction that should appear
> in our published directory.

## Field scope

In scope (as buyer-published): `provider_name` · `licence_number` · `npi` ·
`practice_address`.

**Field contract (R6-7, Tani 2026-09-14T18:08Z):** `npi` denotes the
individual-practitioner NPI (NPPES entity type 1) assigned to a single
practitioner — an organizational NPI (entity type 2) in that field is a ROLE
finding under the ruled field contract (NPI-ROLE-1), never a name mismatch,
and never is an identifier auto-substituted.

**Exclusions (explicit):** network participation, patient acceptance, cost
or quality, taxonomy/specialty, and anything not published by the two named
authorities. A registry field establishes what was PUBLISHED, not what is
true at the front desk.

## Acceptance route

The buyer's reviewer selects any finding and reproduces it from the delivered
package alone — authority evidence, bindings, states — without access to
Word Of Clout's private systems.

## Follow-up — the follow-up cycle (FOLLOW-UP-DEF-1, ruled 2026-09-14T17:09Z)

One bounded cycle after initial delivery, on a **fixed clock**: observations
exactly at **day 7** and **day 30** after delivery — for this engagement,
delivery 2026-09-14, observations 2026-09-21 and 2026-10-14 — the dates
agreed here, in the scope, before work starts. The completion date
(2026-10-14) is bounded; the engagement ends on it.

A notice from the buyer ("we corrected it") is a **trigger or a claim,
never evidence** — and it never restarts the clock.

**Service completion ≠ finding resolution.** Two completed observation
rounds may leave a finding unresolved. `UNCHANGED` and `UNAVAILABLE` are
terminal results of the purchased cycle, **never CLOSED**. The final receipt
can and may say: *service complete, finding unresolved.*

**Coverage:** every agreed remediation candidate from the sampled engagement
is checked; **one retry per unavailable check**; a Word Of Clout method
defect is corrected and rerun **at our expense** and does **not** consume the
retry allowance; when the retry is exhausted, the limitation is reported
explicitly.

**Each result binds:** publication surface, subject, field, expected change
(including an authorized removal), observation time, response evidence +
hash, comparison rule, and result. Observing the agreed publication change
establishes **the publication changed — never the underlying business
fact.** Cycle shape: `follow-up-cycle.jsonl` (schema `woc.follow-up-cycle.v1`),
enforced by `scripts/follow-up-def-gate.mjs`.

## What this engagement does not claim

No directory-accuracy or legal-compliance certification. No participation or
access assertion. Four proposition states only: VERIFIED · UNVERIFIABLE (with
reason) · CONTESTED · STALE — each with its evidence, each signed. The buyer's
own published roster is input as a claim, never proof of its own accuracy.

## Sample

Frozen in `sample-freeze.json` — rules, seeds, and the exact 25 selected IDs,
committed **before any finding ran** (see the freeze statement inside).

## Pricing source

The phase labor ledger — preparation / acquisition / adjudication / delivery /
follow-up — logged as `engagement-labor-ledger.json` during the run. The fee
comes from the measured ledger, not an invented market price.