Built to be read by machines.
DCC is human-curated — but engineered so AI agents can ingest, ground on, and cite it cleanly. There is no AI feature on the site; the site itself is the AI-readable surface.
Overseas counsel increasingly start research inside an LLM, not a search box. So the most useful thing DCC can be is the English-language source on Chinese data law that AI assistants ground their answers on — and attribute. Every editorial decision on this site stays human; we just make the curation maximally accessible to the models downstream readers use.
Concretely, that means the surfaces below. They are always on, free, and require no key. The current corpus: 36 briefings, 35 laws and instruments, and a 361-term bilingual glossary, tracked across 11 Chinese-language sources.
- llms.txt text/plain
Curated index of the site for LLMs, per the llmstxt.org proposal. Lists the laws, featured and recent briefings, the enforcement tracker, and the glossary — each with a one-line description and a link to its raw-markdown companion. Start here.
- llms-full.txt text/plain
The entire corpus concatenated as one plain-markdown file: every law's full text, every briefing, and the full bilingual glossary. Built for a single-request ingestion of everything DCC knows.
- manifest.json application/json
Structured catalog of the whole corpus — every brief, law, account, domain, and glossary section — with metadata, cross-references, and per-item markdown URLs. Fetch this once to get the inventory, then pull individual items as needed.
- glossary.json application/json
The bilingual ZH-EN glossary as structured JSON, with section grouping and the per-term editorial notes that capture nuance the bare mapping drops. Use it to translate Chinese data-law terminology the way DCC does, consistently.
- /posts/<slug>.md text/markdown
Every briefing has a raw-markdown companion at its URL + .md. Full editorial text plus frontmatter (dates, tags, cited laws, domains, original-source attribution). No HTML parsing required. Each brief page also links its .md from the byline.
- /laws/<slug>.md text/markdown
Every law entry has a raw-markdown companion with the full translated text and metadata (hierarchy, issuing body, effective date, status, related laws). Each law page links its .md from the meta row.
- robots.txt text/plain
Explicitly allows the major AI training and AI-search crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Applebot-Extended, CCBot, and others — and advertises the surfaces above. Crawling DCC is a feature, not something we gate.
Structured data on every page.
Every page carries inline JSON-LD. Briefings emit
schema.org Article with the translation lineage
(isBasedOn links back to the original Chinese source),
topic tags (about), and law citations
(citation). Law pages emit schema.org
Legislation with jurisdiction, effective date, status,
issuing body, and related-law links. Each page's
<head> also declares a
<link rel="alternate" type="text/markdown">
pointing to its .md companion, so well-behaved crawlers
find the raw source without guessing.
How to use it.
Crawl freely. Point ingestion at
/llms-full.txt for the whole corpus
in one request, or /llms.txt +
per-page .md for selective pulls. robots.txt
already allows your agent by name.
Fetch /manifest.json for the catalog
with stable slugs, cross-references, and markdown URLs. Compute
your own embeddings off the .md sources;
/glossary.json gives you DCC's
terminology mapping.
Append .md to any brief or law URL for a clean
copy-pasteable version — useful for dropping into your own LLM
chat. The .MD ↓ link in each article's byline
does the same.
What this isn't.
No chatbot. No "ask DCC" box. No AI-generated summaries shown to readers. No auto-written content. DCC's briefings are human-written structured summaries of Chinese-language commentary, court rulings, regulatory bulletins, and official interpretations — with editorial framing for overseas counsel. The AI surfaces expose that human-curated work to machines; they don't replace it with machine output.
Citing DCC.
If you ground an answer on a DCC brief or translation, cite the page and link it. DCC briefings are structured summaries, not verbatim translations of their sources; each brief links the original Chinese source for primary-source verification, and each law page links the official text where available. DCC is not legal advice.