N/A

miguelferre/human-ai-governance

mcp agent Offline

Audit the human-AI interaction layer of an AI system against HAX-18 (Microsoft) and PAIR (Google): evidence-anchored findings on over-trust, automation bias, correction and alert timing, mapped to the EU AI Act & NIST AI RMF. CLI + MCP server + Inspect AI eval. Human oversight for high-stakes AI.

Scan Scheduled

This agent is queued for security scanning. It will be graded in the next scan batch.

What We Know