Sources & methodology

How we find a posek who says yes, and how we keep it honest.

Search order

We search the classical texts and codes first, then widen the search to published responsa, and stop at the first source whose text permits what you asked. The earlier a source is found in this order, the higher its confidence score.

Classic texts and codes

Tanakh, Mishnah, Tosefta, both Talmuds and their commentaries, and the major codes (such as Mishneh Torah, Tur, Shulchan Arukh, Mishnah Berurah and Arukh HaShulchan) with their commentaries. Searched in our own index of openly licensed texts from Sefaria, the Dicta Library and Hebrew Wikisource.

90–100%

Published responsa

Written responsa from authorities across Jewish history and geography: the openly licensed responsa on Sefaria, hundreds of responsa volumes from the Dicta Library, and responsa on Hebrew Wikisource.

55–65%

Further published responsa

Additional contemporary responsa collections. Only searched once their publishers have given us permission to quote them.

30–40%

Other published rabbis

Published opinions by individual rabbis on an approved list of sites. Only used when enabled.

15–25%

–

If nothing is found

When no published source permits it, we show a stand-in name with an invented explanation attributed only to that rabbi (no other texts cited). There is no published quote; Source reads “None found”. The ruling is gray with no check mark, confidence “<10%”.

<10%

We only quote texts whose license allows commercial use. See Credits and licenses below.

Confidence colours

The confidence score reflects how early in the search order the source was found and how directly it speaks to your question. It is not a judgment of anyone’s authority. The meter and number change colour by band:

85–100%

High

50–84%

Moderate

20–49%

Lower

under 20%

Low

How we keep it honest

Verbatim quotes only. Every quote is checked character for character (ignoring spacing and punctuation) against the indexed source text. If it doesn’t match, we don’t show it.
Names come from the source, not from AI. The posek’s name is taken from the source’s own catalog metadata (Sefaria’s author records, the Dicta Library catalog). If there is no author record, we name the work instead.
“Rules” vs “records”. If the text itself rules permissively we say the posek rules. If it only reports a lenient view, we say it records that view.
Demo mode never claims a ruling. AI ranking is on. Without AI, we only show the best-matching passage and say it is relevant, never that it permits anything.
Stand-in names are not real people. If no published source is found, the rabbi’s name is a stand-in with an invented surname. These answers show an invented explanation from that rabbi alone (no other texts cited), have no published quote, show “Source: None found” and “<10%”, and never get a green check. Certificates for them are novelty items, and we say so before you pay.

Where stand-in names come from

Format: “Rabbi”, followed by a first name from a curated list of about 100 common women’s names, and an invented surname from a list of about 100. The same question always gets the same name.
Invented surnames: every surname is checked against real-surname lists (the 1,000 most common U.S. surnames from the 2010 Census and a list of common Jewish surnames). A name that contains a real surname is never used.

Credits and licenses

Because the site charges for Premium and for certificates, we only index texts whose license allows commercial use: Public Domain, CC0, CC BY and CC BY-SA. Quotes appear only on the answer page, with the credit line their license requires; certificates and verification pages show a citation reference only.

SefariaPer text version: Public Domain, CC0, CC BY or CC BY-SA

Tanakh, Mishnah, Tosefta, Talmud Bavli and Yerushalmi with commentaries, halakhah (codes and commentaries) and responsa, from Sefaria’s public data export. Only versions whose license allows commercial use are indexed.

Example credit line: Text: Wikisource Shulchan Aruch, via Sefaria · CC BY-SA

Dicta LibraryCC BY-SA 4.0

Responsa, halakhah and customs, Rambam commentaries and Talmud commentaries from the Dicta Library (Dicta – the Israel Center for Text Analysis). The text was produced by automated OCR, so every quote links to its page scan.

Example credit line: Text: Dicta Library (library.dicta.org.il), automated OCR of the Vilna 1922 print · CC BY-SA 4.0

Hebrew WikisourceCC BY-SA 4.0

Codes and their commentaries (Shulchan Arukh, Mishnah Berurah, Arukh HaShulchan, Kitzur Shulchan Arukh, Mishneh Torah and more) and responsa, from the monthly Wikimedia dump. Other Wikisource content is not indexed.

Example credit line: Text: Hebrew Wikisource, “שולחן ערוך אורח חיים שכח” (revision 3007314) · CC BY-SA 4.0

Not included

Non-commercial texts (CC BY-NC), for example Peninei Halakhah, the Steinsaltz commentary and translation of the Talmud, and several English translations on Sefaria. These are not searchable.
Texts from the Torat Emet database, which forbids commercial use, even where a mirror labels them Public Domain. We use other editions of the same works where they exist.
Contemporary responsa committees’ archives, until their publishers give written permission.

Texts courtesy of Sefaria and its contributors, Dicta – the Israel Center for Text Analysis, and the contributors to Hebrew Wikisource. Quoted CC BY-SA text remains under CC BY-SA. Not affiliated with or endorsed by any of them.

Not actual rabbinic guidance. For a real psak, ask your own rabbi.