Pseudonym MCP is a local pseudonymisation tool designed for privacy-preserving LLM workflows. It detects and replaces PII (personally identifiable information) with opaque tokens before sending text to a cloud-based LLM (like Claude, GPT-4, Gemini), then restores the tokens to the original PII upon return. The system supports multi-language PII detection (English and Polish), hybrid NER (regex and local LLM), strict validation, and is intended as a technical privacy control for GDPR, AI Act, HIPAA, and other privacy frameworks. This makes it ideal for teams and developers integrating LLMs into sensitive data workflows, especially in regulated sectors.
Comprehensive guides on installation, configuration, and usage of Pseudonym MCP.
Community forum for reporting bugs and requesting features related to Pseudonym MCP.
Visit Pseudonym Mcp's official website for product details and getting started.