This MCP server provides a Model Context Protocol interface that lets AI applications and LLMs securely interact with CipherTrust's RESTful Data Protection (CRDP) service. It supports both single-item and bulk protect and reveal operations, with optional versioning support to align with external, internal, or disabled policy configurations defined in CipherTrust. The server routes protect requests to CRDP using policies defined on the CipherTrust manager and returns either the protected data or the corresponding external/internal version values, depending on the policy. Revelation requires proper authorization, either via a username or a JWT token, ensuring that only authorized users can reveal data. Security notes indicate the server runs CRDP in no-tls mode and does not store sensitive data locally, with health and metrics exposed for monitoring. Transport options include stdio (default) and streamable HTTP, configurable via environment variables CRDP_SERVICE_URL, CRDP_PROBES_URL, MCP_TRANSPORT, and MCP_PORT. The project includes integration templates for n8n and supports AI assistant integrations (Cursor AI, Google Gemini, Claude Desktop) for natural-language data protection workflows.