AI Bridge

概要

AI Bridge is a robust and secure Model Context Protocol (MCP) server designed to seamlessly integrate Claude Code with powerful AI models from OpenAI and Google Gemini. It acts as a protective layer, featuring multi-layer input validation, comprehensive content filtering, and sophisticated prompt injection detection to ensure safe and responsible AI interactions. With configurable security levels and flexible model selection, it empowers users to leverage cutting-edge language models while maintaining strict control over data and preventing misuse.

主な機能

• 1 GitHub stars
• Robust error handling with specific error types and detailed messages, alongside Winston-based structured logging.
• Comprehensive security features including multi-layer input validation, content filtering, and prompt injection detection.
• Seamless integration with OpenAI (GPT-4o, GPT-4 Turbo) and Google Gemini (1.5 Pro, 1.5 Flash) APIs.
• Flexible configuration options for AI model selection and inference parameters like temperature.
• Configurable security levels (Basic, Moderate, Strict) and granular controls for all security aspects.

ユースケース

• Securely querying OpenAI models for various tasks directly from Claude Code environments.
• Accessing Google Gemini models for advanced generative AI capabilities with built-in safety controls.
• Developing AI-powered applications within Claude Code that require stringent content moderation and protection against prompt manipulation.