pre-execution-validation
Semantic trust layer that checks AI-agent actions and MCP tool calls before execution.