How does this skill help with the EU AI Act?

It provides templates and implementation patterns for mandatory requirements like Annex IV documentation, Article 12 logging, and Article 14 human oversight.

Can it generate an audit trail?

Yes, it includes a Python-based AuditLogger that creates immutable, chain-hashed records of AI inferences to meet transparency requirements.

Is this skill only for the Dutch government?

While optimized for Dutch standards like BIO2 and the Algoritmekader, its EU AI Act components are applicable to any organization operating under European regulations.

Does it support human-in-the-loop (HITL) workflows?

Yes, it provides structured protocols and decision registration templates to ensure AI outputs are reviewed by humans before being finalized.

GenAI Governance & EU AI Act Compliance

Name: GenAI Governance & EU AI Act Compliance
Author: djimit

bydjimit

•

Data Science & ML

Implements technical governance controls and documentation for Generative AI systems according to the EU AI Act, AVG, and BIO2 standards.

This skill assists developers and government agencies in building compliant Generative AI systems by providing automated templates and code for essential governance requirements. It facilitates the creation of EU AI Act-compliant model cards, audit trails, and human-in-the-loop (HITL) protocols. By integrating technical controls for risk management, transparency, and accountability directly into the development workflow, it ensures that high-risk AI applications meet Dutch government and European regulatory standards.

Key Features

01Provides structured protocols for human-in-the-loop (HITL) oversight

021 GitHub stars

03Automates conformity assessments based on Art. 43 of the AI Act

04Generates EU AI Act Annex IV compliant model cards and technical documentation

05Manages model registries to track versioning, risk levels, and deployment status

06Implements immutable, WORM-compatible audit logging for AI inferences

Use Cases

01Implementing human oversight layers for LLM-generated summaries in administrative processes

02Documenting high-risk AI systems for Dutch municipalities to ensure EU AI Act compliance

03Building tamper-proof audit trails for AI-supported legal decision-making in government

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add djimit/overheid-claude-plugins genai-governance

For use in Claude.ai and ChatGPT

Key Features

01Provides structured protocols for human-in-the-loop (HITL) oversight

021 GitHub stars

03Automates conformity assessments based on Art. 43 of the AI Act

04Generates EU AI Act Annex IV compliant model cards and technical documentation

05Manages model registries to track versioning, risk levels, and deployment status

06Implements immutable, WORM-compatible audit logging for AI inferences

Use Cases

01Implementing human oversight layers for LLM-generated summaries in administrative processes

02Documenting high-risk AI systems for Dutch municipalities to ensure EU AI Act compliance

03Building tamper-proof audit trails for AI-supported legal decision-making in government

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add djimit/overheid-claude-plugins genai-governance

For use in Claude.ai and ChatGPT