190 years of Spanish law as a Git repository.

Ley Abierta reconstructs every Spanish law since 1835 as a Markdown file with a Git commit dated to its official BOE publication: 12,415 laws, 44,553 commits, 18 jurisdictions. Search combines BM25 with int8 vector search on a custom SIMD C kernel, and every model call goes to Qwen on Nan (EU servers, no-log policy). AGPL-3.0. Live at leyabierta.es.

LINKS

leyabierta.es ↗ github.com/leyabierta ↗

STATUS

live

leyabierta.es · agpl-3.0

CORPUS

12,415 laws

44,553 commits · 18 jurisdictions · 1835–2026

INDEX

7.6 GB → 1.4 GB

int8 simd · custom AVX2/NEON · -81%

RETRIEVAL

hybrid

bm25 + vectors · llm rerank · synthesis

INFERENCE

flat-rate

nan · eu servers · no-log · unmetered

THE BRIEF

Public law that's practically inaccessible.

Spanish legislation is nominally public: the BOE exposes every law via an XML API. But the raw feed is a stream of documents with no navigable history. No diff between versions, no graph of which reform amended which article, no way to ask "what changed in labor law in 2018?" and get a useful answer. The official search returns keyword matches in legal prose. Ley Abierta treats the law as source code. Each reform is a commit, each jurisdiction a folder, each version a checkable diff. The result is 190 years of legislation that's queryable and diffable, with full version history.

ARCHITECTURE

BOE to queryable repository in four stages.

  BOE open data API  (XML/JSON · no auth required)
          │
          ▼
  ┌─ pipeline  [ TypeScript · Bun ] ──────────────────────────────────┐
  │   fetch XML → parse → Markdown + YAML frontmatter                 │
  │   git commit --date=BOE_PUBLICATION_DATE                          │
  │   pre-1970 laws: date in frontmatter, git date = 1970-01-02       │
  │   ELI folder structure  (es + 17 CCAA = 18 jurisdictions)        │
  └───────────────────────────────────────────────────────────────────┘
          │                                │
          ▼                                ▼
  leyes/ git repo                   JSON cache (scratch)
  (public human-readable artifact)
          │
          ▼
  ┌─ api  [ Elysia · SQLite · FTS5 ] ────────────────────────────────┐
  │                                                                    │
  │   BM25 full-text search (FTS5)                                    │
  │         +                                                         │
  │   semantic search  — 484K int8 vectors                            │
  │   ┌────────────────────────────────────┐                          │
  │   │  C SIMD int8 cosine kernel          │                          │
  │   │  -81% index size (7.6 GB → 1.4 GB)  │                          │
  │   │  AVX2/FMA + NEON paths              │                          │
  │   │  SharedArrayBuffer worker pool      │                          │
  │   └────────────────────────────────────┘                          │
  │         +                                                         │
  │   RRF fusion  ·  P50 vector stage: 2.1s → 0.8s                    │
  │                                                                    │
  │   Stack: Qwen on Nan (embed + analyzer + rerank + synthesis)      │
  └───────────────────────────────────────────────────────────────────┘
          │
          ▼
  ┌─ web  [ Astro SSG · Cloudflare Pages ] ──────────────────────────┐
  │   law content from leyes/ checkout at build time                  │
  │   derived data fetched from API at build time                     │
  │   live at leyabierta.es                                           │
  └───────────────────────────────────────────────────────────────────┘

  Infrastructure: self-hosted Docker · Watchtower · GitHub Actions cron
                  (daily BOE ingestion) · Resend (email alerts)

DECISIONS

Storage, search, license.

01 · DECISION

Git as the storage layer

Laws are versioned documents. Git is the right tool: git log -- Codigo_Penal.md is a meaningful civic interface, git diff between two commits is a legislative audit trail. Pre-1970 laws are clamped to 1970-01-02 with the real publication date preserved in the YAML frontmatter (Git's commit timestamp is Unix-epoch only).

02 · DECISION

Custom int8 SIMD C kernel

Off-the-shelf vector libraries were too slow at 484K vectors on a single VM. A custom int8 cosine kernel with SIMD intrinsics (dual path AVX2+FMA for x86_64 and NEON for arm64, two-way unrolled accumulator) cut the index from 7.6 GB to 1.4 GB (-81%) and brought the vector-search stage p50 from 2.1s to 0.8s. A SharedArrayBuffer worker pool ensures the index lives once in memory across 4 Bun Workers. No GPU and no managed vector DB.

03 · DECISION

BM25 regressed on citizen queries

BM25 looked obvious for legal text. The early hybrid (BM25 + RRF over Gemini dense embeddings) actually regressed on real citizen queries: FTS5 OR-expansion on "horas extras que no me pagan" matched noise across centuries of legal prose. The fix was a modern-bias prompt on Qwen embeddings that biases retrieval toward recent statutes over 19th-century codes. A reproducible eval harness against citizen and omnibus question sets caught the regression before it reached prod.

04 · DECISION

Full Qwen stack over Gemini + Cohere

The original stack used Gemini embeddings, Gemini Flash Lite as the query analyzer, and Cohere Rerank 4 Pro, all routed through OpenRouter at ~$2 per 1,000 queries with every request leaving the server. A/B evaluation replaced each component with Qwen on Nan, an EU-based inference provider with a no-log policy: embeddings, query analyzer, LLM reranker, and synthesis. Queries stay in the EU, nothing is logged on the inference side.

05 · DECISION

Self-hosted Opik per-span RAG tracing

Every /v1/ask call creates an Opik trace with spans for embed_query, bm25, vector_knn, aggregate_pool, rrf_fusion, rerank, and synthesis. Tracing is fail-safe. Span failures never break the pipeline. Opik runs self-hosted in the same VM (full backend + frontend + Python OSS stack), so per-stage latency is debuggable in prod without exfiltrating queries to a SaaS.

06 · DECISION

AGPL-3.0

Public legal infrastructure shouldn't be capturable by a SaaS. AGPL guarantees that any deployment that serves Ley Abierta over a network must publish its modifications.

OWNERSHIP

One monorepo, sole author.

PACKAGE STACK COMMITS ROLE

pipeline typescript · bun · git — sole

api elysia · sqlite · fts5 · c simd · qwen nan — sole

web astro ssg · cloudflare pages — sole

shared typescript · bun — sole

leyes/ (git repo) markdown · yaml frontmatter · eli folder structure 44,553 sole

OUTCOME

190 years of law, queryable in real time.

Ley Abierta is live at leyabierta.es under AGPL-3.0: 12,415 laws with full version history, and a search layer that holds up on real citizen language without sending queries outside the EU.

NEXT CASE STUDY 05 Asset Reality — withdrawal engine for a law-enforcement custody platform →