Firewall
Createdawilmoth
Filters text data with a powerful rules engine to protect data and enforce policies when using Large Language Models.
About
This tool provides a text filtering firewall service compatible with the Model Control Protocol (MCP). It features a robust rules engine that processes text data before it is sent to Large Language Models, allowing users to identify patterns, enforce policies, and transform text content through customizable rules. Designed for seamless integration with Claude and other MCP-compatible LLMs, it enhances data security and compliance.
Key Features
- REST API for direct integration with applications
- Powerful rules engine with pattern matching and customizable replacements
- Enhanced Docker support for easy deployment
- 1 GitHub stars
- Persistent storage of rules using SQLite
- MCP Protocol Support for integration with Claude and other LLMs
Use Cases
- Enforcing policy compliance through rule-based text transformations
- Customizing text processing for specific LLM applications
- Protecting sensitive information (e.g., SSN, credit card numbers) in LLM interactions