Developers & AI tools

Context bundles (markdown)

One curated, cited markdown file per topic or per law — drop it into your LLM's context, RAG store or project knowledge:

• GET /topics/{slug}/context.md — topic bundle: editorial overview + key law text + case law + enforcement + guidance. Example: /topics/consent/context.md (?lang=nl for Dutch)
• GET /laws/{law}/context.md — full text of a law, cited per provision. Example: /laws/gdpr/context.md — law slugs: gdpr, avg, ai-act, dsa, nis2, uavg, …
• GET /laws/{law}/{article}/context.md — focused per-article bundle: the provision text, related recitals, citing case law, enforcement stats + top fines, guidance and related topics. Example: /laws/gdpr/art-6/context.md
• GET /llms.txt — site summary for AI assistants

MCP server

Connect Claude, or any MCP-capable tool, directly to the corpus. Streamable HTTP, stateless, no auth required (optional Authorization: Bearer olk_… for higher limits):

{
  "mcpServers": {
    "overview-legal": {
      "type": "http",
      "url": "https://overview.legal/api/mcp"
    }
  }
}

Tools exposed:

• search_legal — hybrid search across all content types, results carry the matched passage + topic slugs; mode scopes by jurisdiction (eu default, nl, all), court narrows case law to cjeu or national
• build_overview — CJEU-first, authority-ranked overview of a topic or question (leads with CJEU/EU courts + EDPB, then national case law), with verbatim quotes for the leading judgments
• get_document — one document with topics + citation counts; long judgments/decisions are split into ¶/§ sections — pass query to get the matching passages (quotable verbatim) instead of a top-of-document slice
• get_article — verbatim text of one law article by law + article number (no id guessing)
• verify_quote — anti-hallucination check: confirm a citation (e.g. a C-…/… reference or ECLI) exists, or that a passage appears verbatim
• get_citations — citation graph (law articles ↔ cases ↔ decisions)
• list_topics — find topic slugs by name (EN/NL) to feed get_topic_context
• get_topic_context — the full topic context bundle as markdown
• enforcement_stats — GDPR fine statistics by article, country, authority, year

REST API

curl "https://overview.legal/api/hybrid-search?q=consent+cookies&types=case-law,enforcement&limit=5" \
  -H "Authorization: Bearer olk_YOUR_KEY"

curl -X POST https://overview.legal/api/ask \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer olk_YOUR_KEY" \
  -d '{"q": "When is consent freely given under the GDPR?"}'

API keys & limits

• Anonymous: 60 requests/min per IP on search/citations, 120/min on similar.
• With a free API key: 300 requests/min per key, usage visible in your account.
• Keys are free while the platform grows. Heavy or commercial integrations: mention it via the contact below so we can plan capacity (a paid tier with SLAs will follow demand).

Fair use

• Content is aggregated from official sources (EUR-Lex, curia.europa.eu, rechtspraak.nl, EDPB, national DPAs) and organized by overview.legal; it is legal information, not legal advice.
• Attribute answers and datasets to overview.legal with a link to the canonical URL.
• No bulk scraping of HTML pages — the bundles and API exist so you don't have to.