chore(repo): update GitHub organization references

README.md

@@ -1,8 +1,8 @@
 <img width="400" alt="logo" src="https://github.com/user-attachments/assets/ea105e9e-00d0-4d48-a2a4-006cc4e89848" />
 
-[![CLI](https://img.shields.io/github/v/release/superdoc-dev/docx-corpus?filter=cli-v*&label=cli)](https://github.com/superdoc-dev/docx-corpus/releases)
-[![CDX Filter](https://img.shields.io/github/v/release/superdoc-dev/docx-corpus?filter=cdx-filter-v*&label=cdx-filter)](https://github.com/superdoc-dev/docx-corpus/releases)
-[![codecov](https://codecov.io/gh/superdoc-dev/docx-corpus/graph/badge.svg)](https://codecov.io/gh/superdoc-dev/docx-corpus)
+[![CLI](https://img.shields.io/github/v/release/superdoc/docx-corpus?filter=cli-v*&label=cli)](https://github.com/superdoc/docx-corpus/releases)
+[![CDX Filter](https://img.shields.io/github/v/release/superdoc/docx-corpus?filter=cdx-filter-v*&label=cdx-filter)](https://github.com/superdoc/docx-corpus/releases)
+[![codecov](https://codecov.io/gh/superdoc/docx-corpus/graph/badge.svg)](https://codecov.io/gh/superdoc/docx-corpus)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
 The largest open corpus of classified Word documents. 736K+ real `.docx` files from the public web, classified into 10 document types and 9 topics across 46+ languages.
@@ -12,7 +12,7 @@ The entire document AI research ecosystem runs on scanned images and PDFs. The f
 **[docxcorp.us](https://docxcorp.us)** · **[HuggingFace](https://huggingface.co/datasets/superdoc-dev/docx-corpus)** · **[API](https://api.docxcorp.us/stats)**
 
 > [!IMPORTANT]
-> **WARC trailing bytes fixed for new scrapes** — Documents scraped before [`477d1b9`](https://github.com/superdoc-dev/docx-corpus/commit/477d1b9) may contain a 4-byte WARC record terminator (`\r\n\r\n`) after ZIP EOCD. Future scrapes are fixed; use `python3 scripts/fix_trailing_warc_bytes.py` to inspect or repair older local copies. This was confirmed to trigger Word for Mac's unreadable-content warning for the sample in [#8](https://github.com/superdoc-dev/docx-corpus/issues/8).
+> **WARC trailing bytes fixed for new scrapes** — Documents scraped before [`477d1b9`](https://github.com/superdoc/docx-corpus/commit/477d1b9) may contain a 4-byte WARC record terminator (`\r\n\r\n`) after ZIP EOCD. Future scrapes are fixed; use `python3 scripts/fix_trailing_warc_bytes.py` to inspect or repair older local copies. This was confirmed to trigger Word for Mac's unreadable-content warning for the sample in [#8](https://github.com/superdoc/docx-corpus/issues/8).
 
 ## How It Works
 
@@ -33,7 +33,7 @@ Common Crawl (3B+ URLs/month)
 ## Quick Start
 
 ```bash
-git clone https://github.com/superdoc-dev/docx-corpus.git
+git clone https://github.com/superdoc/docx-corpus.git
 cd docx-corpus
 bun install
 ```

apps/site/content/classification.md

@@ -15,7 +15,7 @@ secondaryQueries:
 
 docx-corpus uses a two-dimensional taxonomy. Every classified document has one document_type label (its form/structure) and one document_topic label (its subject domain). The labels are independent; a `legal` document can have any topic, and a `government` topic can have any document_type.
 
-Taxonomy version: `docx-corpus-v2` (2.0.0). Source: [scripts/classification/taxonomy.json](https://github.com/superdoc-dev/docx-corpus/blob/main/scripts/classification/taxonomy.json).
+Taxonomy version: `docx-corpus-v2` (2.0.0). Source: [scripts/classification/taxonomy.json](https://github.com/superdoc/docx-corpus/blob/main/scripts/classification/taxonomy.json).
 
 ## Document types (10)
 
@@ -114,7 +114,7 @@ See [/quality](/quality) for confidence distributions and the rest of the datase
 
 ## Reproducing this
 
-Code in [scripts/classification/](https://github.com/superdoc-dev/docx-corpus/tree/main/scripts/classification). Run order: `sample.py` -> `label.py` -> `train.py` -> `classify.py`. Pipeline notes in [CLAUDE.md](https://github.com/superdoc-dev/docx-corpus/blob/main/scripts/classification/CLAUDE.md).
+Code in [scripts/classification/](https://github.com/superdoc/docx-corpus/tree/main/scripts/classification). Run order: `sample.py` -> `label.py` -> `train.py` -> `classify.py`. Pipeline notes in [CLAUDE.md](https://github.com/superdoc/docx-corpus/blob/main/scripts/classification/CLAUDE.md).
 
 ---
 

apps/site/content/quality.md

@@ -17,7 +17,7 @@ This page is for researchers deciding whether docx-corpus is fit for their task.
 
 ## Validation
 
-Every file that enters the corpus passes these checks (in [`packages/scraper/validation.ts`](https://github.com/superdoc-dev/docx-corpus/blob/main/packages/scraper/validation.ts)):
+Every file that enters the corpus passes these checks (in [`packages/scraper/validation.ts`](https://github.com/superdoc/docx-corpus/blob/main/packages/scraper/validation.ts)):
 
 1. Minimum size: at least 100 bytes.
 2. ZIP magic bytes: starts with `PK\x03\x04`.

apps/site/src/pages/classification.astro

@@ -47,7 +47,7 @@ const description = 'Two-dimensional taxonomy of 10 document types and 9 topics.
   <section class="prose">
     <div class="section-label">How labels are produced</div>
     <p>We follow the <a href="https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1">FineWeb-Edu</a> pattern: sample a small, stratified set of documents across the five most-common languages, label each with an LLM, then train a lightweight multilingual text classifier on those labels and apply it at scale.</p>
-    <p>The classifier is a fine-tuned XLM-RoBERTa, trained on roughly 3,500 LLM-labeled examples. Two independent classifiers are trained, one per dimension. Code is open: see the <a href="https://github.com/superdoc-dev/docx-corpus/tree/main/scripts/classification">classification scripts</a> on GitHub.</p>
+    <p>The classifier is a fine-tuned XLM-RoBERTa, trained on roughly 3,500 LLM-labeled examples. Two independent classifiers are trained, one per dimension. Code is open: see the <a href="https://github.com/superdoc/docx-corpus/tree/main/scripts/classification">classification scripts</a> on GitHub.</p>
   </section>
 
   <section class="prose">

apps/site/src/pages/index.astro

@@ -863,7 +863,7 @@ const structuredData = {
       <div class="section-label">About the dataset</div>
       <div class="about-prose">
         <p>docx-corpus is the largest open corpus of classified Word documents on the public web. It contains 736,000+ real .docx files collected from Common Crawl, validated, deduplicated, and labeled by document type and topic. It is the first large-scale research dataset for DOCX, the world's most-used document creation format, which has historically been absent from document AI work built on scanned images and PDFs.</p>
-        <p>Each document has a content hash, file size, detected language, and a type and topic label from a fine-tuned XLM-RoBERTa text classifier. The corpus spans 46+ languages and uses a 10-type, 9-topic taxonomy. Available as a <a href="https://huggingface.co/datasets/superdoc-dev/docx-corpus">HuggingFace dataset</a>, a REST API at <a href="https://api.docxcorp.us/stats">api.docxcorp.us</a>, and downloadable manifest files. Built by <a href="https://superdoc.dev/?utm_source=docxcorp.us&amp;utm_medium=referral&amp;utm_campaign=about">SuperDoc</a>. Source on <a href="https://github.com/superdoc-dev/docx-corpus">GitHub</a>. Dataset is <a href="https://opendatacommons.org/licenses/by/1-0/">ODC-BY</a>; source code is MIT.</p>
+        <p>Each document has a content hash, file size, detected language, and a type and topic label from a fine-tuned XLM-RoBERTa text classifier. The corpus spans 46+ languages and uses a 10-type, 9-topic taxonomy. Available as a <a href="https://huggingface.co/datasets/superdoc-dev/docx-corpus">HuggingFace dataset</a>, a REST API at <a href="https://api.docxcorp.us/stats">api.docxcorp.us</a>, and downloadable manifest files. Built by <a href="https://superdoc.dev/?utm_source=docxcorp.us&amp;utm_medium=referral&amp;utm_campaign=about">SuperDoc</a>. Source on <a href="https://github.com/superdoc/docx-corpus">GitHub</a>. Dataset is <a href="https://opendatacommons.org/licenses/by/1-0/">ODC-BY</a>; source code is MIT.</p>
       </div>
       <div class="about-browse">
         <div>

apps/site/src/pages/llms.txt.ts

@@ -40,7 +40,7 @@ ${topics.map(t => `- [${t.label}](${absoluteUrl(routes.topic(t.id))}): ${t.count
 ## Links
 
 - Homepage: ${absoluteUrl(routes.home())}
-- GitHub: https://github.com/superdoc-dev/docx-corpus
+- GitHub: https://github.com/superdoc/docx-corpus
 - HuggingFace: https://huggingface.co/datasets/superdoc-dev/docx-corpus
 - API: https://api.docxcorp.us
 - Takedown requests: help@docxcorp.us

packages/scraper/commoncrawl/warc.ts

@@ -3,7 +3,7 @@ import type { RateLimiter } from "../rate-limiter";
 import type { CdxRecord } from "./cdx-r2";
 
 const WARC_BASE_URL = "https://data.commoncrawl.org";
-const USER_AGENT = "docx-corpus (https://github.com/superdoc-dev/docx-corpus)";
+const USER_AGENT = "docx-corpus (https://github.com/superdoc/docx-corpus)";
 
 export interface WarcResult {
   content: Uint8Array;

scripts/export-hf.py

@@ -127,7 +127,7 @@
 ## Links
 
 - **Website**: [docxcorp.us](https://docxcorp.us)
-- **GitHub**: [superdoc-dev/docx-corpus](https://github.com/superdoc-dev/docx-corpus)
+- **GitHub**: [superdoc/docx-corpus](https://github.com/superdoc/docx-corpus)
 - **Built by**: 🦋 [SuperDoc](https://superdoc.dev)
 """