The zstar MCP server provides a robust and secure gateway for AI agents to interact with sensitive data through the powerful zstar archive utility. It addresses critical security concerns by wrapping battle-tested GPG encryption, zstd compression, and SHA-512 integrity verification into a structured, auditable tool-call API. This ensures that private environment keys, credentials, and sensitive files remain encrypted at rest, are signed for authenticity, and are accessible only to authorized parties, preventing direct shell access risks. It further enhances data protection with features like self-destructing archives for ephemeral data sharing and agent-to-agent encrypted communication, providing complete cryptographic control over every byte an agent touches.
Key Features
01GPG public-key encryption for recipient-specific data access
02Self-destructing archives that securely shred after extraction
03SHA-512 integrity verification to detect data corruption or tampering
04Verifiable GPG signatures for archive authenticity
050 GitHub stars
06Zero shell injection with structured API and parameter validation
Use Cases
01Allowing AI agents to securely manage sensitive data such as API keys and credentials
02Creating auditable and cryptographically controlled archives by AI assistants
03Facilitating secure, encrypted, and signed agent-to-agent communication and data streaming
