MCP Pro is the professional execution-boundary surface for Claude Code, Cursor, Codex, and other supported development tools. It should consume AI SAFE2 Core contracts and return normalized evidence rather than becoming a second policy engine.
The adapter may expose classification, risk scoring, framework mapping, and implementation review according to the deployed product entitlement. The workflow must not infer paid capability from documentation alone.
Integration requirements:
Public reference: https://cyberstrategyinstitute.com/ai-safe2/mcp/