This site was built as a knowledge and analytical atlas based on Markdown files, an internal link network, and editorial and programmatic layers that help transform a broad intellectual corpus into navigable and reviewable reading paths. The site does not rely on a closed editorial database, but on text files that can be inspected and rebuilt, then rendered into static pages and public data files.
This page explains the technical, visual, and methodological resources that went into building the site. Its purpose is transparency: what was used in the interface? what was used in the build? and what are the limits of the tools’ roles compared to human editing? Mentioning any tool or platform here does not mean that it is responsible for the atlas’s content or endorses it; it simply defines where it is used within the workflow.
System Used
The site is based on Quartz, an open-source framework for turning Markdown notes into a browsable knowledge site.
Quartz was modified and customized to suit the nature of this atlas in terms of structure, presentation, Arabic language support, internal links, concept and pathway pages, and source and data indexes.
Pages are built from local Markdown files, then published as static pages that can be browsed and searched.
The customization includes:
- Support for an Arabic reading experience in RTL format.
- Organizing the sidebar according to the atlas structure rather than file order alone.
- Building companion Markdown pages for research use.
- Generating downloadable public data about pages and relationships.
- Linking concept pages, books, paths, and verses through traceable internal links.
Atlas-specific layers that do not come from Quartz alone were also added, including source pages, atom and aggregation pages, conceptual relationship pages, verse-axis pages, and reading and citation tools. Thus, the core framework is Quartz, while the atlas’s structure, method, and links are a custom build on top of it.
Fonts
The site uses the following fonts:
The fonts are used for display and reading purposes only, and in accordance with their respective licenses or terms. They were chosen because they give Arabic text clearer reading space between headings, body text, and technical notes.
The site relies on a functional separation between fonts:
- A font for the interface and short elements.
- A font for long-form body text.
- A font for headings and primary visual markers.
This separation helps keep long research pages readable without turning them into a decorative interface or a purely display-oriented page.
If the fonts cannot be loaded, the site falls back to suitable system fonts so that the pages remain readable. For this reason, access to the content does not depend on loading any particular font, even though the full visual identity appears when the fonts are available.
Tools
Editing and analytical tools were used to organize the material and build the knowledge layers, including Obsidian, local scripts, OCR tools, and language models via OpenAI API, with human review and editing before publication.
These tools enter at different stages:
- Organizing files, links, and preliminary notes.
- Converting material into pages, concepts, and pathways.
- Checking internal consistency and links.
- Building public data for pages and relationships.
- Reviewing drafts and editorial alerts before publication.
Obsidian is used as an editing and organization environment for Markdown files and internal links, not as a publishing platform. Local scripts are used to build indexes and data and to verify the integrity of links and pages before publication.
Language models via OpenAI API are not treated as an independent knowledge source within the atlas, but as assisting tools for analysis, drafting, review, and gap detection. Their outputs are not published as final judgment, nor do they replace recourse to original sources or editorial review.
Distillary is mentioned among the technical and methodological credits as one of the tools or assisting methodologies used to convert some materials into structures that can be indexed, linked, and reviewed. This mention does not make it a source of original content, nor a substitute for Muhammad Shahrur’s books and published materials, nor a party responsible for the editorial conclusions and analyses contained in the atlas.
Data and Scripts
The project uses local scripts to build parts of the site and its data, including:
- Building public page indexes.
- Generating RAG data and inspectable JSONL files.
- Building preliminary relationships between concepts and verses.
- Generating companion Markdown pages for research reading.
- Checking publication integrity before the site is deployed.
These scripts do not change the site’s nature as a static site, but they make the knowledge structure auditable and rebuildable rather than a set of disconnected pages with no data layer.
The build produces a number of lightweight public files within the data section, such as the page index, graph data, and retrieval indicators. These files are not intended to be a full project database, but a technical window that helps inspect what has been built: what pages exist? how are they connected? and which places need review or clearer linking?
More details can be found in Atlas Data, and in Graph Representation for those who want to read the link structure rather than the texts alone.
Sources and Initial Conversion
The atlas relies on returning to original sources when building paths and analyses. OCR, transcription, or preliminary indexing tools may be used to prepare the material, but these outputs are treated as working materials that require review.
Atlas pages do not rely on automated transcription alone in sensitive places, especially book titles, verses, terms, and references. For this reason, there is an editorial layer that separates automated preparation from public publication.
When material appears from a written or audiovisual source, the aim is citation, organization, and analysis, not republication of the source or turning it into a substitute for its original form. For that reason, the rights and use pages are separated from the technical credits page, because the former concerns the limits of using the material, while this page concerns the tools and structure.
Hosting
The site is published via Cloudflare Pages as a static site. Hosting is used to deliver the site and improve performance and access, and it does not change the nature of the content as an independent analytical atlas.
Some supporting Cloudflare services are also used as needed, such as:
- Publishing static pages.
- Managing redirects and links.
- Protecting contact forms when enabled.
- Running some API endpoints related to search or retrieval when needed.
The choice of static deployment is deliberate: the main pages should remain directly loadable and function without full dependence on an interactive application or an administration panel. Dynamic functions, such as contact forms or some retrieval tools, remain an auxiliary layer above the site rather than a condition for reading its pages.
Supporting Services
- Cloudflare Turnstile: for security verification in contact forms, when enabled.
- Google Analytics: for limited general traffic measurement in accordance with Privacy and Cookies Policy.
These services are mentioned because the reader may encounter them in contact forms or in the privacy policy. Their use is limited to operational functions: protection against automated messages, understanding general usage, and improving page accessibility.
Privacy and Measurement
The site uses limited traffic measurement when the reader consents, as explained in Privacy and Cookies Policy.
The purpose of measurement is to understand the most-used pages, improve the reading experience, and detect faults or unclear pathways. These data are not used to alter the atlas’s content according to a reader profile.
Limits of Dependence
The tools mentioned here help with building, but they do not transfer editorial responsibility to any external party.
The atlas’s content, linking choices, pathway formulation, concept ordering, and publication limits remain the responsibility of this project. Technical frameworks and external services assist with presentation, construction, and verification, but they are not a source of content and are not a party to intellectual or copyright endorsement.
As for the rights related to books, audio and video materials, and original sources, they remain with their owners as explained in Copyright and Use Rights.