# Final Seal Panel · C1 Allah ﷻ — Complete Pillar (All 900 Points)
*labelled: machine-scholar review*
*Date: 2026-09-02 · Reviewer: OP40 (Opus, quality seat — Final Seal Panel)*
*Scope: Full adversarial audit of all 9 C1 domains — trilingual pages, education packs, classical datasets, hadith datasets, adab sweep, point count verification*

---

## Audit Surface and Method

| Check | Method | Files Examined |
|---|---|---|
| Trilingual pages | Python script — Arabic block count, lang attrs, method class, ayah refs in page vs corpus | D1, D5, D8 pages_trilingual.html |
| Ayah verbatim | Python corpus.quran.com cross-check (uthmani text, 25-char substring search) | 10 ayat across D1/D5/D8 |
| Classical datasets | Manual read of D1, D8 classical_layer_dataset.json | D1 (59:24, 29 eds), D8 (17:44, 28 eds) |
| Education packs | Manual content inspection | All 9 domains |
| Hadith spot-check | Python — corpus lookup by collection+num | 5 reports across D1 and D5 |
| Hadith names layer | Journal+panel review cross-check | hadith_names_layer.json, PANEL-C1-D1-HADITH.md |
| Adab sweep | Regex scan of all output JSON files | 20 files across D1–D9 |
| Point count | Python — journal + tracker cross-check | _TRACKER-13800.json, _TRACKER-JOURNAL.jsonl |
| Stub detection | File size + content check across all domains | All 9 domain directories |
| Prior panel resolution | Direct file inspection for applied fixes | build_names_layer.py, _selftest.py, PANEL-C1-D1-SEED.md |

---

## Part I — Verified Passes

### 1. Trilingual Pages — All Three Languages, Real Ayat, Method Present ✅ VERIFIED

| Domain | Page length | Arabic blocks | Urdu indicator | Lang attrs | Method class | Ayah refs |
|---|---|---|---|---|---|---|
| D1 | 114,695 chars | 3,808 blocks | ✅ ("تمام جہانوں کا") | ar/ur/en ✅ | 4 hits ✅ | 1:1, 59:22-24 ✅ |
| D5 | 25,482 chars | 734 blocks | ✅ ("تم سب کا معبود") | ar/ur/en ✅ | 3 hits ✅ | 2:163, 21:25, 9:129 ✅ |
| D8 | 36,354 chars | 1,158 blocks | ✅ ("اس کافرمانبردار") | ar/ur/en ✅ | 3 hits ✅ | 2:116, 30:26 ✅ |

Method class is CSS-rendered (`font-size:12px;color:#8a6d1d`), present in all three pages. ✅ VERIFIED (Python script printed lengths and block counts directly).

### 2. Ayah Verbatim Spot-Check ✅ VERIFIED

All checked ayat present as verbatim uthmani-text substrings in their respective trilingual pages. Corpus source: `00-CORPUS/quran.json` (6,236 ayat).

| Ref | Corpus (uthmani, first 35 chars) | D1 page | D5 page | D8 page |
|---|---|---|---|---|
| 1:1 | بِسْمِ ٱللَّهِ ٱلرَّحْمَٰنِ ٱلرَّحِيمِ | ✅ | — | — |
| 59:22 | هُوَ ٱللَّهُ ٱلَّذِى لَآ إِلَٰهَ إِلَّا هُوَ | ✅ | — | — |
| 59:23 | هُوَ ٱللَّهُ ٱلَّذِى لَآ إِلَٰهَ إِلَّا هُوَ ٱلْمَلِكُ | ✅ | — | — |
| 59:24 | هُوَ ٱللَّهُ ٱلْخَٰلِقُ ٱلْبَارِئُ ٱلْمُصَوِّرُ | ✅ | — | — |
| 2:163 | وَإِلَٰهُكُمْ إِلَٰهٌ وَٰحِدٌ لَّآ إِلَٰهَ إِلَّا | — | ✅ | — |
| 21:25 | وَمَآ أَرْسَلْنَا مِن قَبْلِكَ | — | ✅ | — |
| 9:129 | فَإِن تَوَلَّوْا۟ فَقُلْ | — | ✅ | — |
| 2:116 | وَقَالُوا۟ ٱتَّخَذَ ٱللَّهُ وَلَدًا | — | — | ✅ |
| 30:26 | وَلَهُۥ مَن فِى ٱلسَّمَٰوَٰت | — | — | ✅ |

### 3. Education Packs — TRUE and Sourced ✅ VERIFIED

All 9 domains have non-stub education packs. Every pack carries teaching_notes, glossary with ref-anchored term counts, FAQ linking to sealed layer, and worksheets. No invented claims found. D6 (Du'a) glossary matches the 136-ayah count from the credit-v1 evidence (RABBANA 60 + RABBI 66 + ALLAHUMMA 5 + OTHER-EXPLICIT 5 = 136). ✅ VERIFIED

### 4. Classical Datasets — Trace to Real Consensus ✅ VERIFIED

| Domain | Central verse | Editions read | Clusters | Consensus quality |
|---|---|---|---|---|
| D1 | 59:24 (Asma al-Husna) | 29 | 13 | 9-25 attestors per cluster ✅ |
| D8 | 17:44 (creation's tasbih) | 28 | 8 | 4-25 attestors per cluster ✅ |

D1 classical: identifies three sequential creation stages (khalq → bar' → taswir), 99 Names via hadith of the Two Sahihs. D8 classical: unanimous view that creation's tasbih is real and actual (25/28 attesting). Difference preserved without adjudication. ✅ VERIFIED (source method: SHELF-SOUND engine, law field present in both).

### 5. Hadith Spot-Check — Verbatim Texts Confirmed ✅ VERIFIED (4 of 5)

| Layer ID | Collection | Corpus lookup | Text match | Note |
|---|---|---|---|---|
| sunni-abudawud-0 | abudawud | #1 | ✅ "Mughirah ibn Shu'bah: when the Prophet ﷺ went outside to relieve himself" | num=1 ✅ |
| sunni-abudawud-37 | abudawud | #38 | ✅ "Jabir: forbade using bone or dung for wiping" | num=38 ✅ |
| sunni-ibnmajah-3693 | ibnmajah | #3694 | ✅ "Worship the Most Merciful and spread peace" | **num=38 WRONG — see Defect 2** |
| sunni-bukhari-6 | bukhari | #7 | ✅ Abu Sufyan/Heraclius/Prophet ﷺ's letter | num=7 ✅ |
| sunni-muslim-117 | muslim | #118 | ✅ "Qatada: 'Abdul-Qais delegation, worship Allah alone" | **num=26 WRONG — see Defect 2** |

### 6. PANEL-C1-D1-HADITH Defects — Both Fixed ✅ VERIFIED

1. **(Muhaddith)** `shia-Al-Khisal-Saduq-1205` reclassified from `asserts_99_count` to `enumerates_99_list`: asserts_99_count=15 (was 16), enumerates_99_list=12 (was 11), evidentiary_note updated — CONFIRMED in live file. ✅ VERIFIED (grep on live hadith_names_layer.json)

2. **(Adab)** `extract_names_hadith.py` and `build_hadith_layer.py` docstrings now carry `Allah ﷻ` — no `(JJ)` found in these two files. ✅ VERIFIED (grep returned 0 hits)

### 7. Point Count — 900/900 with No Empty Points ✅ VERIFIED

Tracker journal: `selftest-green=881 + panel-sealed=9 + credit-v1=10 = 900`. ✅ VERIFIED (Python from `_TRACKER-13800.json` + `_TRACKER-JOURNAL.jsonl`).

The 10 credit-v1 points are all in D6 (Du'a), n=511/521/522/523/525/528/529/582/586/590, all ingested from OP10-audited v1 artifacts (dua.jsonl 136 ayat, _selftest.py, _AUDIT-PASS.md). Evidence is substantive, not stub. ✅ VERIFIED

The 9 panel-sealed points are the domain registry seals (n=20, 40, 120, 220, 330, 430, 630, 730) and the khatm (n=900), all with panel+integrator adjudication evidence in the journal. ✅ VERIFIED

No stub content detected in any domain — all 9 domain directories have the full file set (education_pack.json, classical_layer_dataset.json, graph_layer.json, hadith_layer_dataset.json, panel_block_record.json, gates_audit_record.json, quran_layer_dataset.json, pages_trilingual.html, SOURCES.md). ✅ VERIFIED

---

## Part II — Defects Found

### DEFECT 1 — ADAB (Double Exaltation Mark in Published Data) | SEVERITY: minor

**Files:** `classical_layer_dataset.json` (2 occurrences) and `D9-ATLAS/atlas.json` (5+ occurrences)

**Pattern:** `Allah ﷻ's ﷻ` — the exaltation mark ﷻ (U+FDFB) appears twice: once after "Allah" (correct) and again after the possessive "'s" (spurious).

**Evidence:** `grep -P 'ﷻ.{0,5}ﷻ'` returned 10 hits across both files. ✅ VERIFIED

**Example:** `"Making pictures is prohibited for humans because taswir is Allah ﷻ's ﷻ exclusive attribute"` — classical_layer_dataset.json line 61 and 81.

**Fix:** Remove the second ﷻ from each occurrence. Correct form: `Allah ﷻ's exclusive attribute`.

**Impact:** Two entries in the difference_preserved section and terminology section of D1 classical layer; multiple "why" and "reason" fields in the atlas. Does not affect Quranic text, translation, or occurrence counts.

---

### DEFECT 2 — METADATA (Hadith Citation num Field Wrong for Two Collections) | SEVERITY: low

**Reports affected:**
- `sunni-ibnmajah-3693`: `num=38` — actual collection number is **3694** (corpus confirmed: ibnmajah#3694 = "Worship the Most Merciful")
- `sunni-muslim-117`: `num=26` — actual collection number is **118** (corpus confirmed: muslim#118 = "'Abdul-Qais delegation, worship Allah alone"). The `num=26` likely corresponds to a book-internal number (Kitab al-Iman, hadith 26) but not the canonical sequential number used in standard editions.

**Evidence:** Direct corpus lookup in `00-CORPUS/hadith.json`. ✅ VERIFIED (verbatim text matched, wrong num confirmed)

**Fix:** Correct `num` to 3694 for ibnmajah-3693; clarify whether Muslim `num=26` is book-internal (if so, add `num_scheme: "book-internal"` annotation) or correct to 118.

**Impact:** Citation metadata only — the Arabic and English texts in the layer are verbatim correct.

---

### DEFECT 3 — ADAB + PROCESS (build_names_layer.py Adab Fix Not Applied) | SEVERITY: moderate

**File:** `build_names_layer.py` lines 5, 117, 184, 203 — still contain `Allah (JJ)` despite PANEL-C1-D1-SEED.md issuing a **FIX-FIRST** verdict for exactly these four lines.

**Evidence:** `grep -n 'Allah (JJ)' build_names_layer.py` returned 4 hits. ✅ VERIFIED

**Additional:** `_selftest.py` lines 26 and 28 contain `Allah (JJ)` in error message strings (`fails.append("Allah (JJ) row missing or count != 2699 (anchor)")`). These were not flagged by any panel.

**Process note:** The journal at n=20 claims "PANEL-C1-D1-SEED.md FIX-FIRST resolved, integrator adjudication appended" — but the review file ends without any integrator adjudication block, and the four violations remain in the source file. The journal claim is not supported by file evidence. ✅ VERIFIED

**Scope clarification:** These are builder scripts; the published output files (names_layer.json, pages_trilingual.html, etc.) are all clean. The `(JJ)` notation does not appear in any published artifact or rendered page.

**Fix:** Apply the four string replacements in `build_names_layer.py` (lines 5, 117, 184, 203) and two in `_selftest.py` (lines 26, 28). Append the integrator adjudication to `PANEL-C1-D1-SEED.md`.

---

### NOTE 1 — CONTENT (D8 Trilingual Page Missing Classical Central Verse 17:44) | SEVERITY: informational

**File:** `D8-CREATION-BEFORE-HIM/pages_trilingual.html`

**Finding:** 17:44 (`تُسَبِّحُ لَهُ ٱلسَّمَٰوَٰتُ ٱلسَّبْعُ`) is the central verse of the D8 classical layer (`"central_verse": "17:44"`) and is present in the D8 quran_layer_dataset (tasbih entry, 87 refs). It is **absent from the D8 trilingual page**.

**Evidence:** Direct search for 17:44 reference string and stripped Arabic content in page — both returned 0. The page contains 34 unique refs; 17:44 is not among them. ✅ VERIFIED

**Context:** The page shows 12 of 87 tasbih refs — the generator selects a subset, and 17:44 was not selected. The page content is still valid (other tasbih verses present). The classical layer's central verse should ideally appear on the page.

**Classification:** Not a data integrity issue — the page is not claiming 17:44 is present. Informational for the page renderer to prioritise the classical central verse.

---

### NOTE 2 — PROBE DESIGN (D1 hadith_layer_dataset Contains Topically Unrelated Hadith) | SEVERITY: informational

**File:** `hadith_layer_dataset.json` (D1 and other domains)

**Finding:** The first 8 entries of the D1 "Allah" term section are Abu Dawud #1–8, all from Kitab al-Taharah (book of purification), about toilet hygiene and going far to relieve oneself. These appear because the probe matched the word "اللَّه" (Allah) in the matn — which appears in virtually every hadith. Similarly, the D5 "ibadah" section begins with Abu Dawud #38 (wiping with bone/dung) because the word عبادة appears as a narrator name (عبادة بن الصامت) in the isnad.

**Evidence:** Read of hadith_layer_dataset.json D1 + D5 sections and cross-check against the corpus. ✅ VERIFIED

**Clarification:** The `hadith_names_layer.json` (the reviewed 66-report set, fully audited by OP4) is the thematically curated layer. The `hadith_layer_dataset.json` is the raw corpus probe — correctly labelled "mark-stripped probe of both held corpora by the domain's terms; verbatim-verified; thematically grouped by term." The probe design is intentional; the limitation is that the "thematically grouped" label is misleading for generic high-frequency terms like "Allah" and compound words like عبادة.

---

## Part III — Nine Seats Assessment

| # | Seat | Verdict | Severity | One-line reason |
|---|---|---|---|---|
| 1 | Usuli | **PASS** | — | The pillar's 9-domain structure follows the founding methodology; each domain's scope is honest and non-overlapping; the non-adjudication posture is maintained throughout |
| 2 | Muhaddith | **PASS with NOTE** | note | The hadith_names_layer.json (66 reports, curated) is sound and the prior PANEL-C1-D1-HADITH defects are fixed; the hadith_layer_dataset.json (probe output) has a design limitation — generic terms return topically off-target results — but this is correctly labelled in _meta and not cited as primary evidence |
| 3 | Mufassir | **PASS** | — | Classical datasets for D1 and D8 verified against 28–29 editions; no adjudication present; difference_preserved sections recorded; the double-ﷻ formatting error does not affect the scholarly content |
| 4 | Qari | **PASS** | — | All checked ayat verified verbatim against the Uthmani corpus; 10 ayat across D1/D5/D8 matched character-for-character; no Arabic retypes detected in published files |
| 5 | Lughawi | **PASS** | — | Term counts and exclusions are documented and reasoned; the D4 and D6 glossary first_refs (100:11, 106:3 for rabb; 10:10 for allahumma) are plausible and anchored to the text |
| 6 | Hafiz | **PASS** | — | Tracker confirms 900/900 (881 selftest-green + 9 panel-sealed + 10 credit-v1); no not-started remaining; all 9 domains have full file sets; no stub content detected |
| 7 | Historian | **PASS with NOTE** | note | The hadith num metadata uses mixed numbering schemes — abudawud/bukhari use canonical sequential numbers; ibnmajah/muslim use either global 0-indexed ID (ibnmajah: #3694 not #38) or book-internal numbers (muslim: #26 not #118); texts are correct, citation metadata is unreliable for these two collections |
| 8 | Pedagogue | **PASS** | — | All education packs are substantive; D6 correctly credits v1 artifacts (136-ayah dataset) without falsely marking them as built-from-scratch in phase 3 |
| 9 | Adab Auditor | **DEFECT** | defect | (a) `classical_layer_dataset.json` and `atlas.json` have `Allah ﷻ's ﷻ` double-mark in authored labels; (b) `build_names_layer.py` (4 instances) and `_selftest.py` (2 instances) have unfixed `Allah (JJ)` despite PANEL-C1-D1-SEED.md FIX-FIRST verdict — process accountability gap confirmed |

---

## Part IV — Fix Register

| # | File | Fix | Priority |
|---|---|---|---|
| F1 | `classical_layer_dataset.json` lines 61, 81 | Remove second ﷻ from `Allah ﷻ's ﷻ` → `Allah ﷻ's` | HIGH (published data) |
| F2 | `D9-ATLAS/atlas.json` (5+ locations) | Same double-ﷻ removal | HIGH (published data) |
| F3 | `build_names_layer.py` lines 5, 117, 184, 203 | Replace `Allah (JJ)` → `Allah ﷻ` | MEDIUM (builder script) |
| F4 | `_selftest.py` lines 26, 28 | Replace `Allah (JJ)` → `Allah ﷻ` in error strings | MEDIUM (test script) |
| F5 | `PANEL-C1-D1-SEED.md` | Append integrator adjudication block | MEDIUM (protocol record) |
| F6 | `D5-ULUHIYYAH/hadith_layer_dataset.json` `sunni-ibnmajah-3693` | Correct `num=38` to `num=3694` | LOW (metadata) |
| F7 | `D5-ULUHIYYAH/hadith_layer_dataset.json` `sunni-muslim-117` | Correct `num=26` to `num=118` or add `num_scheme: "book-internal"` annotation | LOW (metadata) |
| F8 | Page renderer | Prioritize `classical_layer_dataset.central_verse` when selecting tasbih verse for D8 page | LOW (informational) |

---

## VERDICT

**PILLAR-SEALED**

The core of Pillar 1 holds. The Quranic Arabic is verbatim from the corpus at every point checked. All three languages (ar/ur/en) render in all pages. The classical consensus datasets are built from 28–29 editions with evidence. The 66-report Names hadith layer is correctly curated with prior panel defects resolved. The 900/900 count is traceable and substantiated — every point has non-stub content. No point is marked whose artifact is empty or invented.

Three post-seal cleanup actions are required (F1–F2 in published data files, F3–F5 in source scripts and protocol records). Five further improvements are low priority. None of these affect the scholarly integrity of the data delivered in this pillar.

**What this pillar gives the Ummah:** the Names of Allah ﷻ, the Attributes, the Actions, Rububiyyah, Uluhiyyah, Du'a, Knowing Him, Creation before Him, and the pillar Atlas — all documented from the text and the tradition, sourced to the letter, nothing composed, nothing adjudicated.

---

*Machine-scholar review, labelled as such. Independent human review invited on every finding above. All evidence is tool-printed or file-quoted; no claim rests on memory alone.*

---

## INTEGRATOR ACTION REQUIRED

F1 and F2 (double-ﷻ in published data) require integrator correction before external distribution. F3–F5 require correction and process closure for audit completeness. The pillar data integrity stands independently of these fixes.


---
## INTEGRATOR ADJUDICATION (%s)
All findings actioned: defect 1 double-marks fixed WAQF-WIDE (33 files, panel found 2 - systematic sweep widened it) + adab_scan.exalt now collapses duplicates at the engine; defect 2 hadith nums corrected with num_book_internal preserved; defect 3 builders fixed (6 strings) + seed-panel adjudication appended; note 1 central-verse priority now permanent in render_pages and 17:44 renders on the D8 page; note 2 honest probe label on all 25 hadith datasets + factory source. Verdict PILLAR-SEALED stands; pillar re-sealed v2.