FPE Demo
Demonstrates secure format-preserving encryption (FF3 FPE) and multi-mode authentication for sensitive data within LLM interactions.
About
FPE Demo provides a lightweight Model Context Protocol (MCP) server that showcases the practical application of FF3 Format Preserving Encryption alongside versatile multi-mode authentication. It serves as an excellent learning and prototyping tool for understanding how Large Language Models can securely handle sensitive information by encrypting and decrypting data while retaining its original format. Supporting both stdio and HTTP transports, it offers flexibility for local and web-based LLM integrations, making secure data handling concepts accessible and easy to implement in demonstrations.
Key Features
- 0 GitHub stars
- Multiple authentication modes (authless, debug, test, production)
- Provides `fpe_encrypt` and `fpe_decrypt` tools
- FF3 Format Preserving Encryption (radix-10)
- Supports stdio and HTTP transports for LLM integration
- Visually indicates encrypted values with `ENC_FPE:` prefix
Use Cases
- Learning and prototyping secure data handling with LLMs
- Testing web-based LLM connectors (e.g., ChatGPT, Claude) with secure encryption
- Integrating with local MCP-compatible LLM tools like Claude Desktop