Skip to content

Commit 304e39b

Browse files
[docs](consistency): clear stale "pending ratification" wording after the interim ratification
Post-ratification review: ADR-0002's four margins match the script exactly, the empirical-validation follow-up (OI-6) is still tracked, and the D4/D5 values hold up — the only issues were three prose spots still implying the coefficients were unratified: - Model Formulation v0.3: status line and assumption 6 now read "interim-ratified (ADR-0001/0002), pending independent review" instead of "pending Domain-Advisor ratification". - Model Data Sheet: §8 intro and the §9 closing note updated to reflect the new ✅ tier (the v0.3 history row, a dated record, is left as-is). No numeric change; both guard scripts still pass. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
1 parent 833b561 commit 304e39b

2 files changed

Lines changed: 10 additions & 7 deletions

File tree

docs/03-blueprint/model-data-sheet.md

Lines changed: 4 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -300,7 +300,7 @@ These must hold for every input and are unit-tested (NFR-MAINT-2, SRS FR-EDGE-6)
300300

301301
The `qaFit` values are expert heuristics (Charter Section 9), but they are **not invented from
302302
nothing** — each dimension's trade-off shape follows established, widely cited literature, which
303-
is also where a Domain Advisor should start when ratifying them:
303+
underpin the interim ratification in [ADR-0001](../adr/0001-ratify-d4-d5-qafit.md) and are where an independent Domain Advisor should start when reviewing them:
304304

305305
| Dimension | The trade-off shape encoded | Primary sources |
306306
|---|---|---|
@@ -325,5 +325,6 @@ value theory, sensitivity analysis, apportionment).
325325
| C4 Mermaid stub in v1.0? | ❔ Scope | SRS OI-3 |
326326
| Performance budgets ratified | 🧪 Interim set | SRS OI-5 / design DI-4 |
327327

328-
> Numbers in this sheet (🔒 and 🧪) are sufficient to **build and run** the engine today. Ratifying
329-
> the 🧪 values changes only these tables, never the requirements that reference them.
328+
> Numbers in this sheet (🔒, ✅, and 🧪) are sufficient to **build and run** the engine today.
329+
> Reviewing or revising the ✅/🧪 values changes only these tables, never the requirements that
330+
> reference them.

docs/03-blueprint/model-formulation.md

Lines changed: 6 additions & 4 deletions
Original file line numberDiff line numberDiff line change
@@ -5,9 +5,9 @@
55
| Field | Detail |
66
|---|---|
77
| **Document type** | Formal model formulation — the decision model as mathematics, open to expert correction |
8-
| **Version** | 0.2 |
8+
| **Version** | 0.3 |
99
| **Date** | 2026-06-13 |
10-
| **Status** | Baseline — formulation stable; coefficients pending Domain-Advisor ratification (Charter Section 14.4) |
10+
| **Status** | Baseline — formulation stable; coefficients interim-ratified ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md)/[0002](../adr/0002-ratify-preset-calibration.md)), pending independent review (Charter Section 14.4) |
1111
| **Author / Owner** | Faqih Pratama Muhti, B.Sc. Computer Science |
1212
| **Audience** | Software architects, MCDA researchers, reviewers — anyone auditing or challenging the model |
1313
| **Derived from** | [Scoring Algorithm Specification](scoring-algorithm.md) (computation contract) · [Model Data Sheet](model-data-sheet.md) (coefficients) |
@@ -19,6 +19,7 @@
1919
|---|---|---|
2020
| 0.1 | 2026-06-13 | Initial formal formulation: sets, parameters, the pipeline as equations, properties with proof sketches, per-construct literature grounding, and an explicit "how to correct this" section |
2121
| 0.2 | 2026-06-13 | Usability pass: added a by-audience reading guide, a renderer-independent worked example (Section 3.6, = Fixture A), and a note on reading the math in viewers without MathJax; simplified `cases` spacing for robust rendering |
22+
| 0.3 | 2026-06-13 | Status sync: coefficients now interim-ratified ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md), [ADR-0002](../adr/0002-ratify-preset-calibration.md)) — updated the status line and assumption 6 (no longer "pending ratification") |
2223

2324
---
2425

@@ -256,8 +257,9 @@ State openly, so reviewers judge the model on its own declared terms:
256257
5. **Fixed close-call threshold** $\tau = 0.10$ and the equal-weight fallback are conventions, not
257258
derived constants.
258259
6. **Coefficient values.** $A$ and the $\varphi$ vectors (especially the D4/D5 baseline) are expert
259-
estimates pending Domain-Advisor ratification and, ultimately, empirical validation (Charter
260-
Section 11; roadmap v3.0).
260+
estimates — **interim-ratified** by the Owner ([ADR-0001](../adr/0001-ratify-d4-d5-qafit.md),
261+
[ADR-0002](../adr/0002-ratify-preset-calibration.md)), still pending an independent Domain-Advisor
262+
review and, ultimately, empirical validation (Charter Section 11; roadmap v3.0).
261263

262264
## 8. How to challenge or correct this formulation
263265

0 commit comments

Comments
 (0)