docketx/court-rules
US Court Rules — federal and state, held word for word Code & tools: github.com/docketx — legal-scrambler pseudonymises a case file on your own hardware before a frontier model sees it; claude-for-legal is the Claude Code plugin (docketx-open-law) that loads these datasets and checks citations against them. 21,062 court rules across 28 states and the federal courts, in one uniform record. Rules of civil and appellate procedure, evidence, criminal procedure, professional… See the full description on the dataset page: https://huggingface.co/datasets/docketx/court-rules.
US Court Rules — federal and state, held word for word
Code & tools: github.com/docketx — legal-scrambler pseudonymises a case file on your own hardware before a frontier model sees it; claude-for-legal is the Claude Code plugin (docketx-open-law) that loads these datasets and checks citations against them.21,062 court rules across 28 states and the federal courts, in one uniform record. Rules of civil and appellate procedure, evidence, criminal procedure, professional conduct, judicial conduct, general rules and the local rule sets each court publishes for itself.
The text is verbatim. A rule paraphrased is a rule you cannot file against, so every row carries the rule as its court published it, with the URL it came from.
Coverage
Every count is measured from the file in data/ as of 2026-09-11.
The record
Every row is the same shape, whatever court it came from — thirteen columns, all strings:
id, level (federal | state), jurisdiction, court, rule_set, rule_number, citation, title, text (verbatim), source (the URL), license, retrieved_at, extra.
rule_set is whatever name the publishing court uses — a short code (frcp, trcp, wa_rap) where the source is organised that way, the full title ("Delaware Court of Chancery Rules") where it is not. extra is a JSON string (parse it with json.loads) carrying whatever else the source exposed — chapter, article, caption, history note, edition — so that it can differ by court without breaking the shared schema. court is null where the source did not name the promulgating court.
Load it
from datasets import load_dataset
rules = load_dataset("docketx/court-rules") # everything
federal = load_dataset("docketx/court-rules", "federal") # FRCP, FRAP, FRE …
state = load_dataset("docketx/court-rules", "state") # every state setWhat is and is not here
Court rules only. US Code and state statute text live in docketx/us-statutes; 1,712 statute rows were deliberately excluded from this build so that "court rules" means court rules. South Dakota and Montana are the edge case — those two states codify their rules of procedure inside the statute code itself (SDCL ch. 15-6, MCA Title 25), so their rules are cited as statutes but are court rules and are held here. County and municipal local rules are the next layer and are being added jurisdiction by jurisdiction; a state absent from the table above is one whose rules we have not yet verified from an official source, not one whose rules we hold and withheld.
Public-domain government works. Part of the DocketX open legal corpus, alongside docketx/us-caselaw (10.8M opinions), docketx/us-statutes, docketx/oral-arguments-us and docketx/us-judges.