# Briefing for an agent working with the WholeTech / AME material

You have probably been pointed at one page. This lists everything, says what each
thing is, and explains the vocabulary — so you do not have to infer it.

Generated 2026-08-26 06:21. Every figure below is read from the live system at build time.

## Start here, depending on what was asked

| If you were asked about | Read |
|---|---|
| What is outstanding, who owes what | https://wholereach.com/huddle/huddle.md |
| Where the whole thing stands, plainly then technically | https://wholereach.com/state/ |
| Progress against the three questions that matter | https://wholereach.com/scorecard/ |
| Any individual site, scored | https://wholereach.com/alldomains/ |
| Money — spent, ongoing, still needed | https://wholereach.com/finances/ |
| What was said and decided, verbatim | https://wholereach.com/transcript/ame260816-index.md |

## The three questions everything is measured against

These come from Tim Costello, who founded BDX. They are the frame for all of it.

**Two different measurements answer them, and mixing them up is the easiest
mistake to make here.** Both are correct; they measure different things.

| | Scorecard (the six jobs, weighted) | Worksheet (mean of per-site scores, 0–99) |
|---|---|---|
| 1. Works end to end | see /scorecard/ | 56 |
| 2. Phone rang | 0 | 0 |
| 3. Runs unaided | see /scorecard/ | 3 |

The **scorecard** at https://wholereach.com/scorecard/ answers the questions the
way Tim asked them — how much of each of six jobs the department can do. The
**worksheet** at https://wholereach.com/alldomains/ scores every site separately
and averages. If you quote a number, say which one it is.

Question two being zero is not a rendering fault. Nothing in the network has yet
recorded a single tap, form or campaign. It is stated plainly everywhere rather
than softened, and it will only move when money is spent or outreach is sent.

## The network, in numbers

- Domains owned: **426**
- Sites set up in the engine: **263**
- Sites with work published by the engine: **90**
- Sites that could detect an enquiry: **92**
- Knowledge bases built / in use: **78 / 19**

Per-site scores use five axes, each 0–99: **Q1** end to end, **Q2** phone rang,
**Q3** runs unaided, **POT** could marketing plausibly work here, **PRO** is
there money at the end. Network averages: Q1 56, Q2 0, Q3 3, POT 72, PRO 56.
The formula for each is printed on the worksheet itself — if a score looks
wrong, the formula is the thing to argue with.

## Huddle vocabulary

- Items have ids like `h47`. They survive until explicitly closed.
- `carried` = how many published huddle editions the item has appeared in. It is
  **computed** from the editions that exist, not incremented, so rebuilding the
  page cannot inflate it.
- Tags: `ask`, `decision`, `risk`, `shipped`, `idea`.
- `owner` is who owes the next move — commonly Paul, Tim, Melissa or Dawn.

## Caveats worth knowing before you answer a question

- **The transcript is one conversation across twelve days**, 4–16 August 2026,
  not a single day. The `260816` in its filenames is the build date. Earlier
  versions printed only clock times, which made it look like one day where the
  clock ran backwards.
- **The transcript excludes tool output** — that is where credentials and other
  people's personal data live. Redaction runs over what remains.
- **Individual people's names are not published** anywhere in this material. The
  engagement research holds 2,160 real people; only aggregates are ever shown.
- **Scores changed on 16 August** when the scorecard was fixed to compute from
  live data instead of hand-set constants. It previously reported 62/0/20. The
  new, lower numbers are the correct ones.

## Tone that fits this material

Say what is true including when it is unflattering. The zero is stated, the bugs
are recorded, the corrections are kept in the ledger rather than edited out. A
summary that makes it sound better than it is would be the wrong summary.

##### 2026-08-26 06:21 ####