"author:corpussanctum" matching MCP servers:

Enables AI agents to obtain signed permit, deny, or escalation decisions before destructive, deployment, or spending actions, then commit actual effects to a hash-chained, signed audit log.