Skip to content

MCP Server for EU AI Act Compliance Checking - Scan, categorize and verify AI systems against EU AI Act requirements

License

Notifications You must be signed in to change notification settings

desiorac/mcp-eu-ai-act

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

EU AI Act Compliance Checker - MCP Server

CI Version MCP License Python Coverage

The EU AI Act is now in force. Non-compliant AI systems face fines up to 35M or 7% of global revenue. This MCP server scans your codebase, detects AI frameworks, classifies risk level, and tells you exactly what's missing.

3 commands. 2 minutes. Know where you stand.

git clone https://github.com/ark-forge/mcp-eu-ai-act.git
cd mcp-eu-ai-act
python3 example_usage.py

Learn more: arkforge.fr/mcp-eu-ai-act


What It Does

Step Action Output
1 Scan your project Detects OpenAI, Anthropic, HuggingFace, TensorFlow, PyTorch, LangChain
2 Classify risk level Unacceptable / High / Limited / Minimal (per EU AI Act)
3 Check compliance Pass/Fail on each regulatory requirement
4 Report findings JSON report with actionable recommendations

Installation

Via Smithery (Recommended)

smithery install @arkforge/mcp-eu-ai-act

Smithery is the official MCP server registry.

Manual Installation

git clone https://github.com/ark-forge/mcp-eu-ai-act.git
cd mcp-eu-ai-act
python3 server.py

No dependencies required. Pure Python 3.8+.

Quick Start

from server import MCPServer

server = MCPServer()

# Scan a project for AI frameworks
scan = server.handle_request("scan_project", {"project_path": "/your/project"})

# Check compliance (limited risk by default)
check = server.handle_request("check_compliance", {
    "project_path": "/your/project",
    "risk_category": "high"  # or "limited", "minimal", "unacceptable"
})

# Generate a full compliance report
report = server.handle_request("generate_report", {"project_path": "/your/project"})

See examples/ for complete runnable examples.

MCP Tools

scan_project

Scans a project to detect AI model usage.

// Input
{ "project_path": "/path/to/project" }

// Output
{
  "files_scanned": 150,
  "ai_files": [{"file": "src/main.py", "frameworks": ["openai", "langchain"]}],
  "detected_models": {"openai": ["src/main.py"], "langchain": ["src/main.py"]}
}

check_compliance

Verifies EU AI Act compliance for a given risk category.

// Input
{ "project_path": "/path/to/project", "risk_category": "limited" }

// Output
{
  "risk_category": "limited",
  "compliance_status": {"transparency": true, "user_information": true, "content_labeling": false},
  "compliance_score": "2/3",
  "compliance_percentage": 66.7
}

generate_report

Generates a complete compliance report with recommendations.

// Input
{ "project_path": "/path/to/project", "risk_category": "high" }

// Output
{
  "report_date": "2026-02-09T10:30:00",
  "scan_summary": {"files_scanned": 150, "ai_files_detected": 5},
  "compliance_summary": {"risk_category": "high", "compliance_percentage": 50.0},
  "recommendations": ["Missing: Risk management system", "Missing: Human oversight documentation"]
}

EU AI Act - Risk Categories

Category Examples Key Requirements Max Fine
Unacceptable Social scoring, mass biometric surveillance Prohibited 35M / 7% revenue
High Recruitment, credit scoring, law enforcement Full documentation, risk management, human oversight 15M / 3% revenue
Limited Chatbots, recommendation systems, content generation Transparency, user information, content labeling 7.5M / 1.5% revenue
Minimal Spam filters, video games No specific obligations -

Detected Frameworks

  • OpenAI - GPT-3.5, GPT-4, OpenAI API
  • Anthropic - Claude, Anthropic API
  • HuggingFace - Transformers, pipelines, models
  • TensorFlow - Keras, .h5 models
  • PyTorch - .pt, .pth models
  • LangChain - LLM chains, agents

MCP Integration

Works with any MCP-compatible client:

  • Claude Code
  • VS Code with MCP extension
  • CI/CD pipelines (GitHub Actions, GitLab CI)
  • Custom integrations via JSON protocol

Official Resources

Roadmap

  • Integration with EU compliance databases
  • Multi-language support (FR, EN, DE, ES)
  • PDF report export
  • CI/CD native integration (GitHub Actions, GitLab CI)
  • Automatic compliance documentation generation

License

MIT License - See LICENSE for details.


Built by ArkForge | Documentation

About

MCP Server for EU AI Act Compliance Checking - Scan, categorize and verify AI systems against EU AI Act requirements

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •