[realabdu]

What changes

The previous Arabic standard combined naming rules, dictionary field definitions, contribution requirements, and maintenance procedures in one 31-section document. This rewrite makes it easier to review a naming decision and find its examples, while keeping the supporting material accessible to the people who need it.

What is in this PR

• Reorganises both Arabic and English naming standards into 73 rules based on the review sheet, with matching stable three-digit anchors and examples.
• Keeps both standards directly editable as Markdown.
• Restores supporting explanations and developer-facing requirements identified through comparison with the previous document.
• Places dictionary reading and contribution guidance together on the dictionary page, with maintenance procedures in the repository documentation.
• Updates references, generators, packaged documentation, tests, and local preview tooling. Canonical names and the alias index are unchanged; letter-table notes document the identity policy.

1. Explain the naming process before the rules

Restore “Define the concept before naming it” and the convention over configuration introduction. Clarify the document's role: define the concept, choose and form its name using the standard, then record the choice in the dictionary. Rule 001 follows that sequence instead of starting with an unexplained “canonical name”.

2. Keep the rules readable

Add short explanations where they clarify a decision or prevent a likely mistake:

• Why code spelling is predictable and simplified, rather than a reversible scholarly transliteration.
• Why ta marbutah differs between an adjective phrase and an idafah construction.
• Why makki ends in one i, and why letter names such as noon are exceptions to shortened long vowels.
• How tajwid, display name Tajweed, and search spelling tajweed serve different uses.
• Why final hamzah/ayn produces rubu after a consonant but adds nothing in ruku after a long vowel; why popularity alone does not justify a spelling exception.

Retain all 73 rule anchors. Restore the explicit-rule precedence for personal names (shubah, not shuba) and clarify that word references also depend on the text and segmentation method. Rule 038 again distinguishes evidence required before adopted from evidence still incomplete in a draft.

3. Put dictionary guidance on the dictionary page

The dictionary page now contains both how to read an entry and how to contribute to it, in Arabic and English. There is no separate dictionary-contribution page.

• Reader guidance: a combined naming example, field meanings, kind versus category, parent versus part_of, concepts versus registry members, and entry status.
• Contribution guidance: scope and software need, definitions/purpose/boundaries, sources, bilingual consistency, review requirements, and preserving renamed or withdrawn entries.
• Repository setup remains in CONTRIBUTING.md; generation and versioning details remain in tools/README.md.

The dictionary is generated, so the new prose lives in content/fragments/{ar,en}/dictionary-intro.md and is included by the generator. This prevents the next build from overwriting it. Links and the generated skill's dictionary copy are updated accordingly.

Editorial review points addressed

All previously listed points are addressed in commit 4cc302d, in both languages:

• 002–003: retain Arabic based on the concept's specialisation in the Quran or its sciences, regardless of whether an English equivalent exists. General concepts use established technical English.
• 004/007: rule 007 is explicitly an application of 004, keeping its ID for existing references.
• 016: label mawdi as a transliteration-only example; the final general name remains place, matching the data and derivation tool.
• 023: preserve existing letter-name spellings, but require the Arabic letter or codepoint alongside ambiguous names when identity matters. The four cases are shown explicitly. The letter-table notes and decision record now state this policy; dictionary identifiers must still be unambiguous.
• 027/038: define source selection, exact queries, dates, counting units, duplicate/irrelevant-result handling, evidence links and reviewer records. Display choices require relevant evidence; code-spelling exceptions require an explicit term-specific review decision, not an automatic percentage. Historical counts remain preliminary evidence rather than being retroactively certified.
• 044: distinguish linguistic spellings from case/separator-only lookup forms. Both may be recorded in alternative_spellings for resolution, so waqf-lazim remains a valid lookup form without becoming another canonical name. The alias index is unchanged.
• 041: explicitly identify the shortened name as an illustrative counter-example, not a name attributed to a source.

The earlier clarifications to 056 (spelling-rule precedence for personal names) and 069/071/072 (text and segmentation limits on word references) remain included.

The decision record labels these resolutions as proposals in this PR, not upstream adoption. No canonical names were changed and no new usage measurements were fabricated. The full 17-step build and all 173 tests pass locally.

[AbdullahObaid]

Do we need to update the English naming standard now to keep it aligned with the Arabic documentation, even if we use an AI-generated English translation for the time being and review it later? Or would it be better to handle it in a separate PR after this one is finished, so we do not have to keep both versions in sync while the decisions are still being finalized?

[realabdu]

@AbdullahObaid i resolved it . check the PR description for more details