parse-organogram

Extracts Dutch ministerial organogram hierarchies and personnel assignments into JSON proposals.

Updated May 9, 2026
One-click install
npx skills add https://github.com/anneschuth/polder --skill parse-organogram
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: parse-organogram
Source: https://github.com/anneschuth/polder/tree/main/.claude/skills/parse-organogram
Command: npx skills add https://github.com/anneschuth/polder --skill parse-organogram

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Manually extracting hierarchical organizational structure and personnel assignment data from Dutch ministerial organograms is time-consuming, error-prone, and inconsistent, especially when sources are published as PDF images or inline HTML text without standardized formatting.

Core Features & Use Cases

  • Dual Input Support: Process both vision-based PDF/PNG organogram images and text-based inline HTML from ABD manifests, covering all common publication formats for Dutch government organograms.
  • Structured Data Extraction: Generate proposals for both parent-child organizational unit hierarchies (e.g., ministry → DG → Directie → Afdeling) and person-to-post assignments, with confidence scores and evidence citations for every entry.
  • Domain-Specific Validation: Enforce strict Dutch ministerial hierarchy rules, automatically filter out red AVG support roles (policy staff, communicators, jurists), and cap confidence scores for personnel extractions to account for error-prone initialen parsing.
  • Use Case: A public administration analyst building the Polder Dutch government dataset can use this skill to quickly convert a ministry's published organogram into standardized, review-ready proposals without manual data entry.

Quick Start

Use the parse-organogram skill to process the attached 2026 BZK organogram PDF and output the structured organizational hierarchy and personnel assignment proposals.

Frequently Asked Questions about parse-organogram

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract structured hierarchy data from a Dutch government organogram PDF?

To extract structured hierarchy data from a Dutch government organogram PDF, you need a tool that applies vision-based parsing to identify parent-child organizational unit relationships and personnel assignments. This process enforces strict ministerial hierarchy rules and outputs standardized JSON proposals with confidence scores and evidence citations.

Can I parse inline HTML text from an ABD manifest to get organizational unit hierarchies?

Yes, you can parse inline HTML text from an ABD manifest to extract organizational unit hierarchies. The extraction process supports dual input formats, allowing you to process both text-based inline HTML content and vision-based PDF or PNG images to generate structured personnel assignment data.

Does the organogram parsing process filter out specific roles like AVG support staff?

Yes, the organogram parsing process automatically filters out red AVG support roles such as policy staff, communicators, and jurists. It also caps confidence scores for personnel extractions to account for error-prone initialen parsing when generating the final structured data.

What is the best way to convert a published ministry organogram into review-ready JSON proposals?

The best way to convert a published ministry organogram into review-ready JSON proposals is to use an automated extraction tool that standardizes both vision-based and text-based inputs. This enforces Dutch ministerial hierarchy rules and assigns confidence scores for downstream entity resolution.

Are confidence scores assigned to personnel assignments extracted from public administration organograms?

Yes, confidence scores are assigned to personnel assignments extracted from public administration organograms. The extraction mechanism specifically caps confidence scores for personnel data to account for error-prone initialen parsing and provides evidence citations for every entry.