EU AI Act Forcing Questions
This skill pressure-tests any AI system against the EU AI Act before EU deployment, using six Article-cited forcing questions covering prohibited practices, high-risk classification, conformity assessment, role obligations, transparency, and GPAI/systemic-risk status. It produces a structured readiness report with verdicts, citations, and next actions.
✨ What it does
- Checks Article 5 prohibited practices and Article 6/Annex III high-risk classification
- Determines conformity assessment Module A or H per Article 43
- Identifies the organization's role (provider, deployer, importer, distributor) per Article 25
- Verifies Article 50 transparency obligations (AI interaction, synthetic content, deepfakes)
- Assesses GPAI and systemic-risk status under Articles 51-55
- Generates a readiness report with verdict and top 3 actions
🎯 When to use it
- During AI-system intake review for a new system or material change
- Before placing an AI system on the EU market or signing the EU declaration of conformity
- During annual compliance refresh as Article 113 obligations phase in
- When the organization's role changes (e.g., deployer becomes provider via substantial modification)
- When training compute approaches 10^25 FLOPs (systemic-risk threshold)
🚀 How to use
Trigger: /cs:ai-act-readiness
Trigger with the slash command /cs:ai-act-readiness <system> where is the name or identifier of the AI system under review. The skill expects input about the system's characteristics, such as its intended use, data processing, and deployment context. Example prompts:
/cs:ai-act-readiness MyChatbot
/cs:ai-act-readiness facial-recognition-system
📄 Output: A structured EU AI Act readiness report in markdown, including risk classification, conformity assessment, obligation matrix, transparency status, verdict, and top actions.
📦 Add this skill to Claude Code
# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills
# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/compliance-os/skills/ai-act-readiness .claude/skills/ai-act-readinessSkill source: compliance-os/skills/ai-act-readiness/SKILL.md
⚠️ Good to know
Requires accurate system information and may need legal review for ambiguous cases; the skill does not provide legal advice.
❓ FAQ
What happens if the system is classified as prohibited under Article 5?
The system cannot be placed on the EU market; there are no exceptions outside Article 5(2) carve-outs.
When do Article 50 transparency obligations come into force?
Article 50 obligations are in force from 2 August 2025.
What is the systemic-risk threshold for GPAI?
A GPAI is presumed to have systemic risk if its training compute is at least 10^25 FLOPs, per Article 51.
🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.