mcpwall is a security-focused transparent proxy for Model Context Protocol (MCP) servers. It intercepts, logs, and filters all JSON-RPC traffic between your AI coding tools (like Claude Code, Cursor, Windsurf) and MCP servers, enforcing YAML-configured, deterministic, non-AI-based policies to block sensitive file access, dangerous shell commands, and secret leakage. Designed for engineering and security teams managing software agents or LLMs with system-level access, its zero-AI-by-design approach ensures reproducibility and auditability.
Visit mcpwall's official website for product details and getting started.
Comprehensive guides and API references for setting up and configuring mcpwall.
Explore practical examples and use cases for implementing mcpwall effectively.