AgentSkillsCN

process-latvia-data

将 ZVA 注册库中的拉脱维亚国家药品数据整理并归类至各产品文件夹。

SKILL.md
--- frontmatter
name: process-latvia-data
description: Organize and structure Latvian national medicinal product data from the ZVA registry into product folders.

Process Data: Latvian Products

Organize and structure Latvian national medicinal product data from the ZVA registry.

When to Use

When asked to process, organize, or update Latvian product data.

Data Source

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.

Directory Structure

Products are organized under data/latvia/products/ by active substance and brand name:

code
data/latvia/products/<active_substance>/<product_name>/

Folder names are sanitized: lowercased, spaces replaced with underscores, truncated to 96 characters.

Product Folder Contents

Each product folder contains:

  • info.txt - Product metadata with fields:
    • short_name: Brand/product name
    • active_substance: Active substance (Latin name)
    • marketing_authorisation_holder: MAH company and country
    • manufacturer: Manufacturer details
    • atc_code: ATC classification code
    • package_leaflet: Link to patient information leaflet
    • summary_of_product_characteristics: Link to SmPC

Scripts

All 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
  • scripts/combine_latvia_mappings.py - Combines per-product mappings back into the top-level latvia-to-rxnorm.tsv