Hadith Methodology & Provenance

📜The text: which editions

• •The graded collections (Bukhārī, Muslim, Abū Dāwūd, Tirmidhī, Nasāʾī, Ibn Mājah, and more) are served verbatim from verified editions that mirror sunnah.com (the fawazahmed0 Arabic / English / Urdu edition files). The text is displayed exactly as received — never machine-generated, never paraphrased.
• •The breadth library (19 books, ~195k narrations including Musnad Aḥmad, Muṣannaf Ibn Abī Shayba, al-Sunan al-Kubrā and Kanz al-ʿUmmāl) comes from the EasyQuranWaHadees (EQWH) corpus, with full Arabic and Urdu.
• •Editions are served read-only. We add organising layers (search, comparison, tools) around the text — we do not alter the matn.

🔢How numbering works

A citeable reference is only as good as the number being exact. Different collections are cited by different schemes, so each number is shown in the scheme people actually cite — and our canonical link /hadith/collection:number carries that same standard number.

Where no authoritative concordance exists to map a collection onto the international scheme, we keep the honest "printed volume number" label rather than invent a number. An exact, stable number is the whole point of a citeable address — so we never let it drift or guess it.

🏷️Grades: reported, never averaged

• •Every grade is shown WITH the name of the scholar who gave it (e.g. "Al-Albānī: Ṣaḥīḥ", "Zubair ʿAlī Zaʾī: Ḥasan"). Hidayah24 issues no gradings of its own.
• •Grades are NEVER averaged, blended, or turned into a single "% authentic" score. Two scholars may grade the same hadith differently; both verdicts are shown, each attributed — the difference is information, not something to flatten.
• •Bukhārī and Muslim carry an implicit "Ṣaḥīḥ (by consensus)" badge — not a fabricated grader. Weak and fabricated narrations are shown only clearly labelled, and never used to establish belief or law.
• •Alongside the collection’s own grades we also surface a bulk table of grader-attributed gradings (~82,868 rows). Same rule: each row keeps its grader; nothing is merged.

🔍"No source located" is not "fabricated"

Our verification tools are conservative. When a quote cannot be matched to a source we hold, they report "not located" — a statement about our search, not a verdict that the hadith is false. NOT_FOUND ≠ fabricated. Only a scholar’s explicit ruling (mawḍūʿ / lā aṣla lahu) marks a narration fabricated, and that is shown attributed to the scholar who said it.

🧰How verification, takhrīj, isnād & rijāl work

• •Verify a quote — an inline claim-checker that finds the closest verified source and shows a word-level diff. It gives a conservative verdict and, again, treats "not found" as not-found — never as false.
• •Takhrīj — locates a narration across the sources and pulls published sourcing/grading verdicts (e.g. from al-Durar al-Saniyya). Verdicts are never fabricated; when upstream is silent, the tool says so.
• •Isnād explorer & graph — show the chain of narrators as text lists and as a drawn network, so the route of transmission is visible.
• •Rijāl — a ~47,000-narrator biographical directory, so a chain can be read against what is known about each narrator.

🔗The citeable address

Every hadith has a stable, clean permalink in the standard collection:number grammar — for example /hadith/bukhari:1 or /hadith/muslim:2564. It renders the full Hadith 360 view, mirrors the way references are cited elsewhere, and carries the standard citeable number (ʿAbd al-Bāqī for Muslim). The "Cite" button on any hadith copies this permalink together with an attributed citation line.

The older /hadith-360?book=…&n=… links still work — nothing that was ever cited breaks.