Log

2026-04-11

Add Astro site in site/ — single page EN (/) and IT (/it), design system from DESIGN.md
Add GitHub Actions workflow deploy-site.yml (trigger on site/**, uses withastro/action@v3)
Local build verified: 2 pages generated

Add tests/test_get_concepts.py — 10 unit tests for handle_get_concepts tool (subprocess mocking pattern)
- Covers: IT/EN description, not-found, subprocess error, non-zero exit, invalid lang, missing concept_id, default lang, fallback to English, invalid JSON output

Refactor get_concepts into two layers:
- CLI src/istat_mcp_server/cli/get_concepts_cli.py: standalone command istat-get-concepts-cli <concept_id>, JSON output, persistent cache (TTL 1 month)
- MCP tool src/istat_mcp_server/tools/get_concepts.py: wraps CLI via asyncio.create_subprocess_exec, reads name_it/name_en from JSON output
Remove old cli/get_concepts.py (printed all schemes, no concept_id filter)
Register istat-get-concepts-cli as console script in pyproject.toml
Cache key api:conceptschemes:all shared between CLI and MCP tool

Refactor get_data for token optimization:
- No longer calls handle_get_constraints internally — uses get_cached_constraints() directly (no codelist descriptions loaded)
- New _extract_dimension_order(ConstraintInfo) reads Pydantic model directly without JSON roundtrip
- _fetch_parse_filter() caches processed TSV (fetch XML → parse → filter) as a unit — cache hits return ready-to-use TSV
- Cache TTL for observed data updated: 1 hour → 24 hours (86400 seconds)

Add cod_istat, den_rip, cod_rip columns to territorial_subdivisions DuckDB table
cod_istat: numeric ISTAT code for join with TopoJSON (COD_REG, COD_PROV, PRO_COM_T)
den_rip/cod_rip: ripartizione geografica (Nord-ovest, Nord-est, Centro, Sud, Isole)
Mapping derivato da resources/geo/unit_territoriali.csv (NUTS3 2024); match province per nome normalizzato
20/20 regioni e 110/110 province matchano con i TopoJSON
Add scripts/build_admin_boundaries.sh to download ISTAT 2026 admin boundaries (comuni, province, regioni) and produce TopoJSON in resources/geo/
Input CRS EPSG:32632; output GeoJSON 4326 singlepart + TopoJSON fedele (no simplification)
Add get_territorial_codes tool (8th tool) with DuckDB backend
Store territorial data in resources/istat_lookup.duckdb (reusable for future lookup tables)
Include resources/build_territorial_subdivisions.py script to rebuild the DB from ISTAT sources
Add duckdb dependency to pyproject.toml
12 unit tests for territorial codes (level, name, region, province, capoluogo filters)
Update skill docs with territorial codes workflow note