AIR Blackbox MCP Server

EU AI Act compliance scanning for Claude Desktop, Claude Code, Cursor, and any MCP-compatible client.

Unlike other compliance scanners that only report problems, AIR Blackbox also remediates - generating working code fixes, trust layer integrations, GDPR compliance checks, bias analysis, and full compliance reports. Under the hood, the scanning feeds into air-trust, a cryptographic audit chain (HMAC-SHA256) with Ed25519 signed handoffs that ensures compliance data integrity.

14 Tools (10 base + 4 SDK-powered)

Supported Frameworks

LangChain, CrewAI, AutoGen, OpenAI, Haystack, LlamaIndex, Semantic Kernel, Google ADK, Claude Agent SDK, and generic RAG pipelines.

Installation

Recommended (14 tools, every language)

pip install air-blackbox-mcp[full]

Installs the air-blackbox SDK (>=1.17,<2). The scanning tools then run the same engine as the air-blackbox CLI and read Python, Jupyter notebooks, TypeScript, JavaScript, Java and Kotlin, with the same verdict the CLI gives the same code. The floor is the version this package is tested against, and the major cap means a 2.x SDK cannot silently change your findings.

Basic (10 tools, Python only)

pip install air-blackbox-mcp

Works standalone with the lightweight built-in scanner: Python files only, a smaller rule set.

MCP SDK compatibility (mcp 2.0)

This package supports both MCP SDK generations — mcp>=1.0, no upper bound.

mcp 2.0 removed mcp.server.fastmcp and replaced FastMCP with MCPServer. Rather than pin away from it, the server detects which generation is installed and binds to the right class, so it runs on 1.x and 2.x alike:

Both paths are covered by tests that launch python -m air_blackbox_mcp as a real subprocess and drive it over stdio — the same way Claude Desktop and Cursor do — and the full suite runs green on both.

If you are on 0.2.3, upgrade. That version declared an unpinned mcp>=1.0.0, so once mcp 2.0 shipped, every fresh install produced a server that died on import:

ModuleNotFoundError: No module named 'mcp.server.fastmcp'

pip install --upgrade air-blackbox-mcp

0.2.4 fixed it by capping at mcp<2; 0.3.0 removes the cap entirely, so this server no longer conflicts with anything built for mcp 2.x sharing the same environment.

Claude Desktop Setup

Edit ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "air-blackbox": {
      "command": "python3",
      "args": ["-m", "air_blackbox_mcp"]
    }
  }
}

Restart Claude Desktop. The 14 tools will appear automatically.

Claude Code / Cursor Setup

Add to .cursor/mcp.json in your project:

{
  "mcpServers": {
    "air-blackbox": {
      "command": "python3",
      "args": ["-m", "air_blackbox_mcp"]
    }
  }
}

Or add to .claude/mcp.json for Claude Code.

Usage Examples

In Claude Desktop, Claude Code, or Cursor, just ask:

• "Scan this code for EU AI Act compliance"
• "Add a trust layer to this LangChain agent"
• "Check this text for prompt injection"
• "What does Article 12 require?"
• "Generate a compliance report for ~/myproject"
• "Classify the risk level of send_email"
• "Scan this code for GDPR issues" (requires full SDK)
• "Check for bias in this AI model code" (requires full SDK)
• "Can my agent call this shell function?" (requires full SDK)
• "Show me my compliance trends" (requires full SDK)

SDK Features (Optional)

The full air-blackbox SDK unlocks 4 additional tools:

1. GDPR Scanning (scan_gdpr)

   • Personal data handling without consent
   • Data retention and erasure policies
   • Cross-border transfer safeguards
   • Data processing agreements

2. Bias Analysis (scan_bias)

   • Disparate impact risk detection
   • Protected attribute handling
   • Training data bias indicators
   • Fairness metric awareness

3. Action Validation (validate_action)

   • Pre-execution approval gates (Article 14)
   • ConsentGate policy enforcement
   • Risk-based action filtering
   • Audit trail generation

4. Compliance History (compliance_history)

   • Track past scan results
   • Analyze compliance trends
   • Export audit trails
   • Monitor improvement over time

Optional: Deep Analysis with Ollama

For AI-powered analysis beyond regex patterns:

# Install Ollama
brew install ollama

# Pull the fine-tuned compliance model
ollama pull air-compliance-v2

# The analyze_with_model tool will automatically use it

What Makes This Different

Other MCP compliance tools only scan. AIR Blackbox:

1. Scans + Remediates - finds issues across 6 EU AI Act articles AND generates working code fixes
2. Analyzes deeply - regex patterns + AI-powered model analysis + prompt injection detection (15 patterns)
3. Validates before execution - pre-approval gates and risk classification for agent actions (Article 14)
4. Tracks compliance - GDPR checks, bias analysis, full reports, and historical trend monitoring (SDK)

Architecture

Which engine runs is decided once, at import, by what is installed, and every result says which one ran. Until 0.4.0 the scanning tools always used the built-in scanner even when the SDK was installed; the two had drifted (fewer checks here, Python only, and a hiring-detection bug the SDK had fixed), so the same code got different verdicts from the CLI and from Claude.

The choice is made when the server starts, not per call, and provenance.engine names it on every result, so two reports are comparable exactly when their engine and ruleset_version match.

Result provenance

Because a compliance finding is only comparable to another if you know what produced it, every machine-readable result carries a provenance block:

{
  "findings": [ ... ],
  "provenance": {
    "engine": "builtin-rules",
    "scanner_version": "0.2.4",
    "ruleset_id": "eu-ai-act-art9-15",
    "ruleset_version": "cee71577c486",
    "sdk_version": null
  }
}

• engine — builtin-rules or air-blackbox-sdk, whichever actually ran.
• ruleset_version — a content hash of the active rules, not a hand-maintained string. Change a regex and it changes by itself; a version someone must remember to bump is one that eventually misreports which rules ran.
• sdk_version — the SDK that produced this result, so it is null for built-in results even when the SDK is installed alongside. Reporting a version that contributed nothing would imply its rules ran.

Two reports with the same engine + ruleset_version were produced by byte-identical rules and can be diffed directly. Different values mean the rules moved, and the diff needs that context to be meaningful.

Errors carry provenance too — knowing which version produced an error is as useful as knowing which version produced a finding.

Install [full] to unlock the Tier 5 SDK tools; the base install works standalone.

Part of AIR Blackbox

This MCP server is part of the AIR Blackbox ecosystem:

• air-trust on PyPI - the cryptographic audit chain that backs compliance scanning
• air-blackbox on PyPI - the full compliance SDK and CLI scanner
• airblackbox.ai - the project homepage and docs

Links

• EU AI Act - the regulation
• GDPR - data protection regulation