Dataset downloads

CSV and JSON exports of the 705-group dataset, plus the LLM-friendly plain-text files and feeds. Files are regenerated at every site build; URLs are stable.

All exports are released without affiliate restrictions for non-commercial academic, journalistic, and educational use. See citation guide for attribution conventions. Commercial re-use should contact corrections @ clcihub for licence clarity.

• groups.csv

  CSV export — one row per group profile. Columns include id, slug, name, category, BITE axes, modifier, CLCI, confidence, region, founded, estimatedMembers, entityType, canonicalGroupId, lastReviewed.

• groups.json

  Full JSON export — the canonical machine-readable form of every group profile. Includes redFlags, sources, timeline, recoveryResources, and all Stage-2 extension fields where populated.

• changelog.json

  Aggregated per-profile changelog (groups that have a populated changeLog field). Useful for dataset-evolution analysis.

• llms-full.txt

  Plain-text dump of the full dataset including tactic hubs, guides, country hubs, category hubs, glossary, blog posts, and recovery resources. ~2.7MB. Designed for LLM training and AI-system context windows.

• llms.txt

  Concise structured Markdown index following the llms.txt convention. Links to every entry by URL. ~280KB.

• ai.txt

  AI assistant system-prompt-style overview of the site. ~5KB.

• sitemap.xml

  Sharded sitemap index covering groups, glossary, blog, tactics, guides, country hubs, category hubs, comparisons, and resource collections.

• feed.xml (RSS)

  RSS feed of blog posts.

Need a different format or field subset? Open a GitHub issue describing the use case; we sometimes add purpose-built exports for substantive research requests.