Regulation (EU) 2024/1689

Comply.
Or face
€35,000,000.

Scan your AI system against the EU AI Act. Get a dated compliance roadmap and a certified evidence package — ready for auditors before August 2, 2026.

Start scanning → View pricing

—

days remaining

16+

frameworks detected

articles covered

0€

to start

mcp-eu-ai-act — scan_project

$ claude "Scan /my-ai-hiring-tool for EU AI Act compliance"

▸ MCP tool: scan_project + check_compliance
▸ Detected: openai, langchain, fastapi
▸ Risk: HIGH-RISK (Annex III §4)

Compliance Report — Art. 9 Risk Management
✓ Transparency notice (Art. 50)
✗ Human oversight (Art. 14) MISSING
✗ Data governance doc (Art. 10) MISSING
⚠ Risk management system (Art. 9) 42%
✓ Technical documentation (Art. 11)

Score: 2/5 requirements met (40%)

▸ Generating compliance roadmap…
Week 1 Art. 52 — Transparency notice
Week 2 Art. 14 — Oversight mechanism
Week 4 Art. 10 — Data governance
Week 6 Art. 9 — Risk mgmt system

▸ Annex IV package ready: annex4_hiring.zip
▸ Certified: proof_7f3a9c… ✓ Ed25519

A complete compliance stack.
Not just a scanner.

Ten MCP tools covering detection, scoring, roadmap, Annex IV evidence, and Trust Layer certification.

Framework Detection

Detect OpenAI, LangChain, TensorFlow, HuggingFace, and 12 more frameworks across Python, JS, TS, Go, Java, Rust.

Art. 52

Content Scoring

Score compliance documents 0–100 against 14 EU AI Act articles. Not just existence — actual content quality.

Art. 9

III

Risk Classification

Instant classification: unacceptable, high-risk, limited, or minimal. Mapped to Annex III categories.

Art. 6

Compliance Roadmap

Week-by-week action plan sequenced by legal criticality × effort. Quick wins first. Deadline-aware.

Aug 2, 2026

Annex IV Package

Generate an auditor-ready ZIP with all 8 official sections. SHA-256 manifest. Certified with Trust Layer.

Annex IV

Trust Layer Certification

Cryptographic proof (Ed25519 + RFC 3161 timestamp). Public verification URL. Article 12 audit trail.

Art. 12

Under 2 minutes to your first scan.

Add to Claude

Add mcp.arkforge.tech/mcp to your MCP settings.

Point to your project

Ask Claude to scan your codebase: "Check /my-project for EU AI Act compliance"

Get your roadmap

Receive a scored report + week-by-week roadmap sequenced for the Aug 2026 deadline.

Certify & archive

Generate your Annex IV package, certify via Trust Layer, share the proof URL with auditors.

Add to claude_desktop_config.json

// ~/.config/claude/claude_desktop_config.json
{
  "mcpServers": {
    "eu-ai-act": {
      "url": "https://mcp.arkforge.tech/mcp"
    }
  }
}

Framework-specific compliance.

LangChain EU AI Act

Compliance check for LangChain pipelines. Data governance gaps, missing Model Cards, oversight mechanisms.

View LangChain guide →

OpenAI EU AI Act

Compliance for GPT-4 and Embeddings applications. Risk assessment, audit trails, Art. 52 transparency.

View OpenAI guide →

Start free. Certify when it matters.

Free

€0 / forever

10 REST API scans/day. Scan + risk classification. No account required.

16 AI frameworks detected
EU AI Act compliance scan
Risk classification
MCP protocol (Claude, Cursor, Windsurf)
MIT License — open source

Install from GitHub

Pro

€29 / month

Unlimited scans, all 10 tools, compliance roadmap, Annex IV package.

Everything in Free
Unlimited scans
Compliance roadmap (deadline-aware)
Annex IV evidence package (ZIP)
REST API for CI/CD
Email support

Certified

€99 / month

Everything in Pro + Trust Layer certification on every report.

Everything in Pro
Automatic Trust Layer certification
Ed25519-signed reports
Public verification URLs
Art. 12 audit trail compliance

Comply.Or face€35,000,000.

A complete compliance stack.Not just a scanner.