# 07 — Gaps, open questions and honest tensions

---

## 1. Build-fresh items

| # | Item | Why |
|---|---|---|
| **G1** | **The naming ruling** (`03__` §4) | The corpus flagged the collision and left it unresolved. The site cannot use the word until this page exists |
| **G2** | **A second working twin** | One (S3) proves the interface claim; the pattern claim needs two. The AWS IAM twin is closest — finish and publish it |
| **G3** | **The coverage measure, computed once** | Provisions-without-twins over the AI Act is fully specified and never run. One number would make `/discipline/` concrete — and it feeds `standards.sgit.ai` directly |
| **G4** | **The 2FA demo** | The designed milestone for actors + fractal build + change pipeline. Everything downstream labels itself against whether this exists |
| **G5** | **A twin schema** | *"properties, behaviours, functions, inputs and outputs"* is stated everywhere and typed nowhere. One `Schema__Twin` in the house style would move the definition from prose to code |
| **G6** | **The name-watch check** | Grep the estate for new "twin" usages per build (`04__`). Cheap, and the collision already happened once |

## 2. Open questions

| # | Question | Where it stands |
|---|---|---|
| **Q1** | **How connected is a twin, really?** The agent-twin brief itself carries *"the standing question of how connected the twin actually is to reality"* | Connected-or-not is binary in the model; freshness, latency and fidelity are not. A twin updated *"manually once a week"* is connected — and six days stale. The discipline needs a staleness dimension |
| **Q2** | **Who may update a twin?** Twins act, and twins are updated — but nothing says who is authorised to change a twin's definition, which is now the graph's contact with reality | The execution-broker composition (`03__` §4.3) suggests the answer; nothing states it |
| **Q3** | **Does the discipline of reality survive simulation?** The world model runs counterfactuals — attacks that have not happened. But *"everything has to exist"* | The likely answer: simulated states are marked as projections from real twins, never facts. Nobody has written it |
| **Q4** | **Is the broker's concentration risk acceptable?** It *"inverts the catastrophic failure property the rest of the architecture depends on"* | The corpus states the cost and does not decide. This is a real architectural fork for the estate |
| **Q5** | **When is a twin done?** *"even one specific use case can become massive once you capture all the elements"* — the same brief offers no stopping rule | The grounding ladder's stopping test (*"the last node where going deeper would neither improve observability nor change a decision"*) probably applies; say so |

## 3. Honest tensions

1. **One working twin, nine applications.** The primitive is load-bearing across five sites and demonstrated once. Publishing that ratio is the credible move; the S3 container carries a lot of weight until G2.
2. **The discipline of reality vs the simulation ambition.** The cluster's most distinctive rule and its most ambitious design pull in opposite directions, and Q3 is where they meet.
3. **"Connected or not" is the powerful claim and the oversimplified one.** Binary connectedness makes coverage computable; Q1's staleness dimension is what it hides.
4. **The broker inverts the estate's core property, on purpose.** Publishing that trade as a decision-to-be-made is more honest than presenting the Service Twin as an unqualified design.
5. **The best material arrived last and least verified.** The 19 August pair is 8,488 words, converted from an external session, with market research "as supplied." Strong ideas, weakest provenance in the cluster — label accordingly.

## 4. Loose ends worth an hour each

- Verify the S3 container's current state in the sg-compute repo (the vault specs shipped; confirm the S3-face specifically).
- Check whether the "digital twins already built" the AWS engine reuses are on disk, and where.
- Run the twin name-grep once now — there may already be a third meaning in the corpus.
- Confirm the 2FA scenario's public framing keeps clear of the Feb GRC records (`05__` §2).
- Ask whether the Service Twin market research's vendor claims were ever checked.

---

This document is released under the Creative Commons Attribution 4.0 International licence (CC BY 4.0).