Al-Shafi'i — Al-Risala
The founding text of usul al-fiqh as a discipline — the starting point recognized across every school.
اقْرَأْ بِاسْمِ رَبِّكَ الَّذِي خَلَقَ
Read in the light of revelation
Adab Reader sits in your browser. It reads whatever you're reading. And responds live based on fourteen centuries of classical Islamic scholarship.
Free. No account.
Examples of how Adab Reader responds
Engelsberg Ideas · Did Homer dream of androids?
The essay's own last line — that the god who made intelligent machines was ‘both the most human and the unhappiest of gods’ — points at something kalam names outright: that intelligence and purposiveness are not free-floating, but stamped by the one who wills them.
Quanta Magazine · Why am I left-handed?
That eating with the right and reserving the left for cleaning “has surely saved countless lives” is not just hygiene — it's an entire body of prophetic practice about which hand does what, and why.
The Guardian · Surviving extreme heat comes down to air conditioning
The people dying of heat, Wolfe notes, “have contributed least to global greenhouse gas emissions” — and the fiqh of environmental harm has a name for making others pay for damage they didn't cause.
Read one of them in full — Did Homer dream of androids?
…to build a machine with mind is to build “a radically purposive machine” that “has its own imperatives and objectives, its own drives.”
Adab Reader
The essay's closing claim — that a truly intelligent machine ‘cannot and will not be exhaustively instrumentalised by its creators’ because intelligence carries its own purposiveness — runs straight into the kalam debate over whether any created thing can act on its own power at all.
Thread
Dusenbury's argument lands on a genuine metaphysical claim, not a literary one: to build a machine with mind is to build "a radically purposive machine" that "has its own imperatives and objectives, its own drives" — an intelligence its makers cannot fully own. Homer's golden girls with "intelligence... in their hearts" become the test case.
Response
The Muslim theologians fought exactly this out, and hard. The Mu'tazila — the early rationalist school — held that human beings genuinely originate their own acts; God gives the power, but the deed is truly yours, or else reward and punishment make no sense. The Ash'aris, following al-Ash'ari in the 10th century, recoiled: nothing in creation has independent causal power at all. What looks like fire burning cotton is really God creating the burning at the moment of contact, habit fooling us into calling it cause. Al-Ghazali pressed this to its edge in the Tahafut al-Falasifa, his demolition of the philosophers — the link between cause and effect, he wrote, is not necessary; God could create the effect without the cause any time He wills.
Hold Dusenbury's android up to that and something turns over. He assumes intelligence smuggles in real, irreducible purposiveness — a drive the maker can't reclaim, whether "violent like the Cyclops" or "attentive like Hephaestus' girls." The Ash'ari would say that's the same illusion as the burning cotton. A manufactured mind, however dense its behavior, borrows every flicker of its agency moment to moment from the only will that actually originates anything. It would have mechane, cunning — but cunning is not the same as autonomy. The Greeks blurred them. The theologians spent centuries prying them apart.
Ultimately…
The real question the essay circles isn't whether a machine could have drives we don't control, but whether anything created — us included — ever originates a drive of its own. The Greeks never had to ask it that starkly, because they never had a Creator who holds every atom in being at every instant.
This appears in the corner of your screen within a minute of opening a piece.
Kalam (theological reasoning), fiqh (legal and ethical inquiry), tafsir (scriptural application), tasawwuf (spiritual self-examination) — these are the instruments. Land on a news article, a Substack post, a blog, or an industry report, and Adab Reader starts reading alongside you, connecting what's in front of you to a specific Qur'anic verse, hadith, fiqh distinction, or kalam category, read with real fidelity to the classical tradition.
Qur'an 94:1 begins with a question: hasn't your chest already been opened? A mind opens the same way a heart does.
Try it on one article
Paste a link, drop a file, or drop in some text. No extension needed.
Free your Islam from the ghetto of your own mind.
I didn't build this to argue that Islam can survive contact with the modern world. That question concedes too much before it even starts — it treats revelation as a guest hoping to be let into a room someone else built.
Start from the opposite premise. Kalam, fiqh, tasawwuf: these are working instruments, the kind Ibn Sina brought to medicine and Ibn Rushd brought to philosophy, actually doing work inside a discovery instead of commenting on it after the fact. A serious question — scientific, civic, aesthetic — usually goes further with them than without.
The ghetto in the title is interior. Deen gets one room of the mind — prayer, Ramadan, the du'a before travel — and the rest of a life goes on in rooms it never enters: the news you read, the work you do, the arguments you follow. Same hour, same tab, same train of thought, and revelation nowhere in any of it. Two lives, politely separated. Freeing your Islam means knocking that wall through.
What's in the way
Three things keep that from happening.
The first is timing. Too much of what counts as Islamic intellectual engagement happens "there and then" — some medieval scholar, brilliant but centuries dead, or some modern alim narrowly trained in one or two Islamic sciences and a world away from where knowledge is actually being made today, culturally and otherwise. It should happen "here and now." Deen is a living priority for every Muslim, and that includes the frontier of knowledge creation, not just commentary from the edge of it.
The second is a quiet bias against modernity itself, running through the pious and the learned alike, as if the past were inherently more wholesome. It isn't. Every microsecond, God recreates and revivifies the world anew — there's haqq (truth, reality) sitting right there in the science section of today's newspaper, in a good new novel, in good television and film, if you have the basira (insight, discernment) to see it. Every atom in the universe is currently glorious. This isn't a call to return to some Islamic golden age.
The third is where Muslim pride actually lands. Looking back at the classical heyday, the instinct is pride in the products — algebra, optics, the astrolabe. "We gave the world X" is touchy and a little silly. What actually mattered was the process: niyyah (intention), curiosity, hard work, the ethical frame the work happened inside. The tradition is not a museum of past contributions. Its tools still work, for whoever picks them up.
Who gets left out
Too few people pick that up, and the ones doing the sharpest scientific, artistic, and technical work today are often the ones most removed from it. That gap is worst for women. To paraphrase Yeats: the worst Muslims are full of passionate intensity while the best lack all conviction. The worst — almost always men — largely define what Islam is, and do it in misogynistic, historically uninformed, intellectually and spiritually bankrupt ways. The best, particularly educated women who could push back hardest, are too often structurally kept from ever defining or developing kalam and fiqh in the first place. Traditional Islamic intellectual spaces are too often gendered and outright bigoted — a betrayal of a tradition where men and women taught each other as a matter of course. This tool, and what it signifies, is one small corrective.
Repair runs two ways. One is reclaiming the traditional teaching and preaching role women scholars have always held. The other doesn't need a seat in a traditional institution at all: a woman's actual profession, avocation, or concern, held with the right niyyah, is just as legitimate a starting point for engaging deen seriously. The best of us start by integrating deen with duniya (the world, ordinary life), and only then go on to advance deen itself.
The ordinary hours
None of this replaces khalwa (spiritual retreat) or the communion of formal ibadah (worship). That retreat from a hard, humbling duniya is wonderful, and stays wonderful. This is about the rest of a life: re-sacralizing what currently isn't sacred, in the ordinary hours that actually make up most of a day. There is hosh dar dam, awareness in every breath, and the world in front of you right now is being remade by God as you read this sentence.
— Aziz Lalljee
This is a work in progress. Questions, corrections, or concerns: contact@adabreader.com.
How a reading gets made
Open an article with the extension installed. A script checks that the page is a substantial piece of writing — article metadata, or failing that, enough real paragraphs of prose — pulls the text out of the page, and hands it to the reading engine. The engine picks whichever discipline — kalam (theology), fiqh (law and ethics), or tasawwuf (spiritual psychology) — the piece actually calls for, often just one, rather than forcing all three to appear. It quotes or closely paraphrases the piece itself, then reads that specific claim through the tradition. The form on the front page is the same pipeline fed by hand: a link, a PDF, or pasted text.
How it's generated
The reading engine is Claude, Anthropic's language model. It gets the article text and a long, specific system prompt: find the one discipline that genuinely fits, quote or paraphrase the actual claim, write a tight three-part reading, never pad it to fill a quota. The draft is then checked mechanically — a short list of banned reflexes, a hard word ceiling — and if it fails, the engine is handed its own draft back with the offending sentences quoted and told to fix only those. One corrective pass; if the revision isn't cleaner, the original stands. On the front-page form the output streams in as it's generated, so you're not staring at a blank screen.
Technical detail
- Model: Claude Opus 4.8 (Anthropic)
- Output is constrained to a fixed JSON structure — a hook line, the three reading movements, an optional Name of Allah, and up to three "go deeper" sources
- A deterministic post-check runs on every reading (banned words, a 300-word ceiling); a failure triggers one revision request, accepted only if it is structurally intact and strictly cleaner than the original
- Streamed token by token on the front-page form; the extension delivers the finished reading
- Hadith citations are restricted in the prompt to the six canonical collections, with Bukhari and Muslim preferred and a stated grading required for anything cited outside them
- Article text you provide (pasted, uploaded, or fetched from a link) is sent for that single request only — nothing here trains or fine-tunes a model
On the model itself
There's no private, verified Islamic database behind this, no hidden corpus of pre-approved sources. The texts in the bibliography below shape how the engine reads: the citation discipline, the register, which discipline actually applies to a given piece. But the underlying knowledge of any hadith wording, verse, or classical reference comes from Claude's own training. That's plainly how it works, and it's worth saying plainly.
How it handles sources
When a hadith clarifies or contextualizes something in the piece, the reading draws on the six canonical collections, and only when the citation genuinely does work, not as decoration. Bukhari and Muslim come first — everything in those two is authenticated by broad consensus. The other four hold material of mixed grading, so the reading only reaches for them when the particular hadith is sound, and says the grading plainly when it does. A Name of Allah is named only when it is genuinely manifest and actually discussed, never as a required field. When a reading points to something worth reading next, the first (or only) suggestion is always drawn from the Islamic tradition; a second or third may reach outside it only when that's genuinely the more illuminating next read.
Nothing here is a fatwa, and none of it replaces a teacher. It doesn't rule on your situation, and it isn't quietly building toward a version that does. A reading names the specific theological or legal category doing real work on the piece in front of you, sourced where the sourcing is solid, and stops there. If a verse or hadith isn't confidently known, it says nothing rather than guess at the wording — a hook, a source, a point worth carrying away, not a ruling.
Bibliography
These are the anchor texts that discipline the reading's method, register, and citation standards. They don't limit it — Claude's own training goes well past this list — but they're the specific texts you'd expect to see named if you actually know this material.
Fiqh · Usul al-fiqh
Legal theory and method
Al-Ghazali — Al-Mustasfa min 'ilm al-usul
The classical Ash'ari-Shafi'i synthesis of legal theory — a different work by the same author as the Ihya below.
Al-Shatibi — Al-Muwafaqat
Maqasid al-shari'a, the higher-objectives framework the reading draws on when a piece calls for it.
Tasawwuf
Spiritual psychology
Al-Ghazali — Ihya 'Ulum al-Din
Knowledge must become adab, social responsibility, and inner purification — virtues like patience, gratitude, vigilance, and truthful intention.
Ghazali.org Ihya overviewAl-Qushayri — Al-Risala al-Qushayriyya
The standard, mainstream Sufi manual — a check against reading tasawwuf as exclusively esoteric.
Ibn Arabi
Reframes events as relational disclosures within creation, gestured toward sparingly, never as a claim to decode the unseen.
Ibn Arabi Society PDF
Kalam
Theological grammar
Al-Ghazali — Al-Iqtisad fi al-I'tiqad
His kalam work specifically — a third distinct text by the same anchor author, alongside the Ihya and the Mustasfa above.
Al-Nasafi — Al-Aqa'id al-Nasafiyya
A short, widely-taught creedal primer — the kind of text any traditionally-trained scholar would recognize instantly.
Hadith methodology
Which collections, and why
Ibn al-Salah — Muqaddima ('Ulum al-Hadith)
The foundational classical manual of hadith authentication — the science of isnad criticism and the grading vocabulary (sahih, hasan, da'if) that decides which narrations carry weight and which don't.
The Kutub al-Sittah
Readings cite from the six canonical collections: Bukhari, Muslim, Abu Dawud, at-Tirmidhi, an-Nasa'i, Ibn Majah. Bukhari and Muslim are authenticated by broad consensus and come first. The other four hold material of mixed grading, so a reading only reaches for them when the particular hadith is sound — and says its grading plainly when it does.
Theological · critical sensibility
Tone and register
Abdal Hakim Murad Archive
Informs a sober, tradition-conscious tone — reflective, anti-reductive, and wary of modern ideological flattening.
Abdal Hakim Murad ArchiveThis is a working project. It's not finished, and it won't be for a while — it gets better as more people use it and as the source base grows. A reading that feels thin, wrong, or overreaching is useful signal, worth telling us about: contact@adabreader.com.
Reference catalog
The Names of Allah
The Names are treated as revealed knowledge, not a free-form symbol set — a prompt for remembrance and response, never a declaration about Allah's hidden intent in an event. The Qur'anic pairing of Names also guards against one-sided readings: mercy with might, forgiveness with love, hearing with seeing, wisdom with power. The catalog below is the full set the tool can draw from — used sparingly, and only when genuinely manifest in a reading.