菜单
Organize and structure Latvian national medicinal product data from the ZVA registry into product folders.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Search for RxNorm standard concepts using the Hecate semantic search API. Primary search tool for drug mapping.
Parse EMA Authorised Presentations PDF files into structured TSV format. Use when asked to parse, extract, or process raw EMA product PDFs, or to fetch/download new EMA data.
Map Spanish national medicinal products from the AEMPS CIMA registry to RxNorm standard concepts. Use when asked to map a Spanish product, review or correct a Spain `mapping.tsv`, validate a Spain product folder, fix a bad EMA auto-link via `nro_definitivo`, or handle Spain-specific cases such as duplicate `nro_definitivo`, generics, biosimilars, herbal products, or radiopharmaceuticals.
General principles for mapping pharmaceutical products to RxNorm concepts. Covers mapping types (EXACT/BROAD/NO_MAPPING), terminology adaptation, unit conversions, and common pitfalls.
Map Latvian national medicinal products to RxNorm standard concepts. Use when asked to map a Latvian product or create/review a mapping.tsv in a Latvia product folder.
Map EMA centrally authorised product presentations to RxNorm standard concepts. Use when asked to map an EMA product or create/review a mapping.tsv in an EMA product folder.
| name | process-latvia-data |
| description | Organize and structure Latvian national medicinal product data from the ZVA registry into product folders. |
Organize and structure Latvian national medicinal product data from the ZVA registry.
When asked to process, organize, or update Latvian product data.
Latvian State Agency of Medicines (ZVA) registry: https://dati.zva.gov.lv/zalu-registrs/export/en
Raw data is stored as data/latvia/HumanProducts.json.
Products are organized under data/latvia/products/ by active substance and brand name:
data/latvia/products/<active_substance>/<product_name>/
Folder names are sanitized: lowercased, spaces replaced with underscores, truncated to 96 characters.
Each product folder contains:
info.txt - Product metadata with fields:
short_name: Brand/product nameactive_substance: Active substance (Latin name)marketing_authorisation_holder: MAH company and countrymanufacturer: Manufacturer detailsatc_code: ATC classification codepackage_leaflet: Link to patient information leafletsummary_of_product_characteristics: Link to SmPCAll scripts are located in the scripts/ subdirectory of this skill and use git-root-based paths. They can be invoked from any directory in the project.
scripts/organize_products.py - Creates the folder structure from HumanProducts.json, generates info.txt files/workspace/scripts/generate_mapping_overviews.py latvia - Regenerates latvia-to-rxnorm.tsv from per-product mappings (generic script, not skill-specific)