Open-source compliance infrastructure for AI agents — built for the EU AI Act.
The August 2, 2026 enforcement deadline for high-risk AI systems is approaching. AIR Blackbox provides the audit trails, security controls, and compliance tooling that enterprises need to deploy AI agents legally in the EU.
Every AI agent decision gets recorded, every tool call gets gated, every prompt gets scanned — with tamper-evident HMAC-SHA256 audit chains that regulators can verify.
pip install air-compliance # Scan your project for EU AI Act gaps
air-compliance scan ./my-project # Get a compliance report instantlyDrop-in compliance for every major AI agent framework:
| Package | Framework | Install |
|---|---|---|
| air-langchain-trust | LangChain / LangGraph | pip install air-langchain-trust |
| air-crewai-trust | CrewAI | pip install air-crewai-trust |
| air-openai-agents-trust | OpenAI Agents SDK | pip install air-openai-agents-trust |
| air-autogen-trust | AutoGen / AG2 | pip install air-autogen-trust |
| air-rag-trust | RAG Knowledge Bases | pip install air-rag-trust |
| openclaw-air-trust | TypeScript / Node.js | npm install openclaw-air-trust |
| Package | Purpose | Install |
|---|---|---|
| air-compliance | EU AI Act compliance scanner | pip install air-compliance |
| Gateway | Flight recorder reverse proxy | docker pull ghcr.io/airblackbox/gateway:main |
| air-platform | Full stack (one command) | docker compose up |
Each trust layer provides controls mapped to specific articles:
- Article 9 — Risk Management: ConsentGate classifies tool calls by risk level
- Article 10 — Data Governance: DataVault tokenizes PII, ProvenanceTracker hashes KB documents
- Article 11 — Technical Documentation: Structured audit logging with call graphs
- Article 12 — Record-Keeping: HMAC-SHA256 tamper-evident chains
- Article 14 — Human Oversight: Exception-based blocking, audit trail review
- Article 15 — Robustness: InjectionDetector, WriteGate, DriftDetector
Core Runtime: Gateway, Agent-Episode-Store, Agent-Policy-Engine, Air-Platform
Instrumentation: Python-SDK, air-langchain-trust, air-crewai-trust, air-autogen-trust, air-openai-agents-trust, openclaw-air-trust, air-rag-trust
Safety & Governance: OTel-Collector-GenAI, OTel-Prompt-Vault, OTel-Semantic-Normalizer, Agent-Tool-Sandbox, Runtime-AIBOM-Emitter, AIBOM-Policy-Engine
Evaluation & Testing: Eval-Harness, Trace-Regression-Harness, Agent-VCR
Security: MCP-Security-Scanner, MCP-Policy-Gateway
Compliance: Air-Compliance-Checker
Apache 2.0 · Built on OpenTelemetry · PyPI
