Hadith is a science of provenance.
Every hadith comes with a chain of named people, each one examined, and fourteen centuries of scholarship on whether that chain holds. The Hadith Project is putting all of it online: open, linked, and checkable in seconds.
Not a new collection, and not a new grading. A verifiable record of what the scholars established, with the sources, built in the open.
Quoting takes a second. Checking takes a library.
Hadith are quoted, disputed and dismissed millions of times a day. The texts sit in dozens of editions with incompatible numbering. The narrators' records sit in biographical dictionaries that fill shelves. The verdicts are scattered through a thousand years of books. So the arguing happens without the evidence, and a method of rare rigour looks, from the outside, like a matter of faith.
None of this is lost. It is simply not linked. Linking it is an engineering problem, and engineering problems get solved.
- 2,602
- connected hadith, without repetition, that al-Bukhari admitted into his Sahih by Ibn Hajar's count, from the six hundred thousand reports he said he sifted.
- 12 × 12
- ranks of reliability, and generations, into which Ibn Hajar sorted every narrator of the six canonical collections and their appended works in the Taqrib.
- 5
- conditions a report must satisfy before anyone may call it sahih.
What it takes to call a hadith sahih
The categories are not a mood. They are a decision procedure, fixed in writing by Ibn al-Salah in the seventh century from the practice of the critics before him. A report earns the word sahih only when all five conditions hold.
- An unbroken chainاتصال السند
Every narrator heard from the one before, all the way back to the source.
A chain of custody with no gap in which the evidence could have been swapped.
- Upright narratorsعدالة الرواة
Each person in the chain was known for honesty and religious integrity.
A witness whose character was vetted by people who knew them.
- Precise narratorsضبط الرواة
Each transmitted accurately, from memory or from a trustworthy written copy.
A witness whose recall was tested against everyone else's.
- No anomalyعدم الشذوذ
The report does not contradict what more reliable narrators transmitted.
A result that is not an outlier against stronger data.
- No hidden defectعدم العلة
Experts who compared every route found no subtle flaw.
Peer review by the people who have read every other copy.
Fail any one of the five and the report is weak (da'if). Keep all five with a narrator of slightly lesser precision and it is hasan. A chain with a light weakness can be strengthened by independent chains saying the same thing; a liar's chain cannot. Weak never means false; it means unproven by this route. How the science works
A chain, examined
This is what every hadith page will look like. Switch the depth, click a narrator, and read what the critics wrote about them. The chain below is the first hadith in Sahih al-Bukhari.
Actions are by intentions
Adds the chain, each narrator's record, and the scholars' verdicts.
إِنَّمَا الأَعْمَالُ بِالنِّيَّاتِ، وَإِنَّمَا لِكُلِّ امْرِئٍ مَا نَوَى، فَمَنْ كَانَتْ هِجْرَتُهُ إِلَى دُنْيَا يُصِيبُهَا أَوْ إِلَى امْرَأَةٍ يَنْكِحُهَا فَهِجْرَتُهُ إِلَى مَا هَاجَرَ إِلَيْهِ
Actions are only by intentions, and each person will have only what they intended. So whoever migrated for worldly gain, or for a woman to marry, their migration was for whatever they migrated to.
The chain, person by person
- Reliable
- Truthful, lesser precision
- Weak
- Abandoned
- Unassessed
- Companion
Read upward from the book to the Prophet ﷺ. Click a narrator to see the record critics kept on them.
View this chain as a table
| Narrator | Standing | Died | Received it with |
|---|---|---|---|
| The Prophet Muhammad ﷺ | The Prophet ﷺ | 11 AH | source |
| Umar ibn al-Khattab | Companion | 23 AH | I heard the Messenger of God ﷺ say (سمعتُ رسول الله ﷺ يقول) |
| Alqama ibn Waqqas al-Laythi | Reliable | died during the caliphate of Abd al-Malik (65–86 AH) | I heard Umar, on the pulpit (سمعتُ عمر على المنبر) |
| Muhammad ibn Ibrahim al-Taymi | Reliable | 120 AH | that he heard Alqama say (أنه سمع علقمة يقول) |
| Yahya ibn Sa'id al-Ansari | Reliable | 144 AH | informed me (أخبرني) |
| Sufyan ibn Uyaynah | Reliable | 198 AH | narrated to us (حدثنا) |
| Abdullah ibn al-Zubayr al-Humaydi | Reliable | 219 AH | narrated to us (حدثنا) |
| Muhammad ibn Isma'il al-Bukhari | Reliable | 256 AH | narrated to us (حدثنا) |
Select a narrator to open their record: dates, teachers, and what the critics wrote, with the source for each line.
- · A mountain of memory, and the world's leading authority on the understanding of hadith.
- · Reliable, a hadith master and jurist; the most eminent of Ibn Uyaynah's students.
- · Reliable, a hadith master, jurist, imam and proof; however his memory changed late in life, and he would sometimes transmit with an ambiguous formula, though only from reliable narrators.
- · Reliable and firm.
- · Reliable; has some reports no one else transmits.
- · Reliable and firm, of the second generation. Whoever claimed he was a Companion erred; it is said he was born in the Prophet's lifetime.
- · Companion
Who graded it, and why
We do not grade hadith. Each line is a named scholar's verdict, where they gave it, and the reason they gave. Disagreement is shown, not resolved.
| Scholar | Verdict | Where | Reason given |
|---|---|---|---|
al-Bukhari d. 256 AH | Sahihصحيح | Sahih al-Bukhari 1, 54, 2529, 3898, 5070, 6689, 6953 | Chose it to open the Sahih and recorded it in seven places through five of Yahya ibn Sa'id's students (Ibn Uyaynah, Malik, al-Thawri, Hammad ibn Zayd, Abd al-Wahhab al-Thaqafi). Inclusion in the Sahih is itself a grade: every narrator meets his conditions. |
Muslim ibn al-Hajjaj d. 261 AH | Sahihصحيح | Sahih Muslim 1907 | Included in the Sahih, in the Book of Leadership. |
al-Tirmidhi d. 279 AH | Hasan sahihصحيح | Jami' al-Tirmidhi 1647 | Sound, and known only through Yahya ibn Sa'id. He also records Ibn Mahdi's remark that this hadith ought to be placed in every chapter. |
al-Nawawi d. 676 AH | Sound by consensusصحيح | al-Arba'in, hadith 1; al-Minhaj Sharh Sahih Muslim 13/53 | Opened his Forty with it, noting that the two imams of hadith recorded it in their Sahihs, the soundest books composed. In his commentary on Muslim he writes that the Muslims agree on its great standing, its many benefits and its soundness. |
Ibn Rajab al-Hanbali d. 795 AH | Agreed uponصحيح | Jami' al-Ulum wa al-Hikam, hadith 1 | Scholars agreed on its soundness and received it with acceptance; al-Bukhari opened his Sahih with it in place of a preface. Records al-Shafi'i's remark that it enters seventy chapters of law and Ahmad's that it is one of three hadith on which Islam's foundations rest. |
Ibn Hajar al-Asqalani d. 852 AH | Sahih, agreed uponصحيح | Fath al-Bari, commentary on hadith 1 | Agreed upon by all; recorded by every famous imam except in the Muwatta. Answers al-Tabari's observation that on some people's method it would be rejected for being single-stranded: it is single-stranded only in its sound form, and only in this wording; the meaning is attested from other Companions, and those who called it mass-transmitted erred. |
Ibn Taymiyya d. 728 AH | Among the rare reports of the Sahihصحيح | Majmu' al-Fatawa, vol. 18 | Sound and single-stranded at once: his phrase for exactly this shape of chain. |
What this chain teaches about the method
At its first four links this is a single-strand report (gharib): one Companion, one Successor, one student, one student again. It is still one of the most universally accepted hadith in the corpus, because every one of those people is of the highest grade. Rarity is a property the discipline records; weakness is a verdict it reaches only from the narrators.
“Narrated to us” and “I heard” assert direct hearing. Sufyan ibn Uyaynah occasionally used an ambiguous “from”, and critics noted it. Here, in the standard recension, he says “narrated to us”, which closes the question before it is asked.
A report that Sufyan's memory failed in his last year is on record alongside his rank as an imam, and so is al-Dhahabi's rejection of that report. Ahmad's reservation about Muhammad ibn Ibrahim sits beside four critics who called him reliable. The record hides nothing; it lets a reader see how the balance was struck.
Below Yahya ibn Sa'id the chain widens into hundreds of routes. A report can be rare at its root and widely attested at its trunk; both facts are tracked, and both matter for how later scholars weigh it.
Terms used here
- Isnad إسناد
- The chain of people who passed the report down, named in order.
- Matn متن
- The text of the report itself.
- Gharib غريب
- A report carried by a single narrator at some point in its chain.
- Mashhur مشهور
- A report carried by three or more narrators at every level, short of mass transmission. This hadith is rare at its root and became mashhur from Yahya ibn Sa'id onward.
- Thiqa ثقة
- A narrator critics judged both upright and precise.
- Tadlis تدليس
- Using an ambiguous formula like “from” that hides whether the narrator heard directly.
- Muttafaq alayh متفق عليه
- Recorded by both al-Bukhari and Muslim from the same Companion.
Two more worked examples, one genuinely disputed and one weak, are in Explore. The weak one is a line almost everyone has heard.
One link that ends the argument
Paste a Hadith Project link anywhere and it unfurls into the text, the source, the chain and who graded it. The person you are talking to does not have to trust you. They can look.
Behind the link is a stable identifier. A cite names the collection and a declared canonical numbering, with a published crosswalk to every other edition. A fingerprint of the normalised Arabic text lets anyone check that a quotation matches the record, whichever edition it came from.
- Cite
- thp:bukhari:1
- Fingerprint
- sha256:6fd00f9ec66e
- Same data, for machines
- /api/v0/hadith/actions-by-intentions.json
إِنَّمَا الأَعْمَالُ بِالنِّيَّاتِ، وَإِنَّمَا لِكُلِّ امْرِئٍ مَا نَوَى، فَمَنْ كَانَتْ هِجْرَتُهُ إِلَى دُنْيَا يُصِيبُهَا أَوْ إِلَى امْرَأَةٍ يَنْكِحُهَا فَهِجْرَتُهُ إِلَى مَا هَاجَرَ إِلَيْهِ
Actions are only by intentions, and each person will have only what they intended. So whoever migrated for worldly gain, or for a woman to marry, their migration was for whatever they migrated to.
- al-Bukhari
- al-Humaydi
- Sufyan ibn Uyaynah
- Yahya ibn Sa'id al-Ansari
- Muhammad ibn Ibrahim al-Taymi
- Alqama ibn Waqqas
- Umar ibn al-Khattab
- the Prophet ﷺ
https://thehadithproject.com/h/actions-by-intentions/Mission
To make the sciences of hadith as transparent, searchable and verifiable as the best modern reference works, by building open data, open tools and open standards that scholars trust and anyone can use.
Vision
A world in which no hadith is quoted without its chain, no chain without its narrators, and no verdict without its reasons. Where checking a hadith takes seconds, misquoting one is harder than quoting it correctly, and the method of the muhaddithin is recognised for what it is: one of the most rigorous systems of source criticism people have ever built.
Goals
- One stable reference for every narration prototyped
A universal identifier with a published crosswalk across editions and numberings, so a cite means the same thing everywhere.
- Every narrator on record prototyped
A structured, sourced entry for every narrator of the major collections, quoting the critics in their own words.
- Every chain drawn prototyped
Interactive isnad graphs for every narration, with transmission formulae and continuity visible.
- Every verdict with its reason prototyped
Gradings from classical and contemporary scholars side by side, each tied to its source, with disagreement shown rather than averaged away.
- Three depths, one record in the prototype
Reader, student and researcher views of the same page, so a newcomer and a specialist are looking at the same facts.
- Open by default prototyped
Open data, open code, a public API and embeddable reference cards, so apps, bots and browser extensions can cite instantly.
- Governed by scholars planned
An editorial process in which qualified hadith scholars review every change, and nothing automated is ever presented as a verdict.
What we will not do
Trust is the whole product. These are the lines that keep it.
- We will not grade hadith.
Scholars grade. We show who said what, where they said it, and the reason they gave.
- We will not hide disagreement.
Where al-Albani and al-Arna'ut differ, both appear, with their reasoning. Where the classical critics split, the split is the record.
- We will not publish a claim without a source.
Every line traces to a book, an edition and a page. Where a scan exists, it links to the scan.
- We will not let an algorithm pose as a muhaddith.
Search, text clustering and continuity checks surface questions for human review. They never answer them.
- We will not lock the data.
Everything is downloadable, licensed for reuse, versioned, and mirrored. If this project ends, the data outlives it.
The road
Prototype
- Three hadith with every link examined: one sahih, one disputed, one weak
- Identifier scheme and text fingerprint
- Reader, student and researcher depths
- Reference cards and a static JSON API
- The blueprint, open for criticism
Foundation
- The six canonical collections, with a numbering crosswalk across editions
- A record for every narrator of the six collections, built from the Tahdhib literature
- Verdicts from the classical critics and from al-Albani and al-Arna'ut, with reasons
- Isnad parsing for every narration, reviewed by people
- An editorial board of hadith scholars and a public change log
Infrastructure
- A browser extension and reply bot that answer any cited hadith with its card
- Search by meaning, labelled as retrieval, never as verdict
- Manuscripts and scanned editions linked at page level
- Further collections: Musnad Ahmad, the Muwatta, Ibn Hibban, al-Hakim and beyond
- Partnerships with universities and hadith institutes
Build it with us
The scholarship exists. The data exists, in pieces. What is missing is the work of joining them with care, and that is work many hands can share.
- Hadith scholars
Review narrator records and verdicts, write the reasons in plain language, and sit on the editorial board.
- Developers
Isnad parsing, narrator disambiguation, the graph store, the API, the extension and the bot.
- Translators and editors
Make the records read well in English, Urdu, Turkish, Malay, French and beyond.
- Institutions and funders
Lend authority, host mirrors, and fund the scholars who do the reviewing.