This is a security proxy that wraps your MCP servers to monitor and enforce policies on every tool call in real time. It intercepts prompts, file operations, and MCP requests, then sends redacted payloads to a cloud classifier that checks for data leaks, PII exposure, and prompt injection before returning allow, redact, or block decisions. Primarily designed to work through VS Code and Cursor extensions rather than as a standalone MCP server, though the Python proxy can be integrated directly. Reach for this if you're running AI coding agents with access to sensitive codebases or customer data and need semantic analysis beyond traditional DLP tools that only see destinations, not intent.
claude mcp add --transport stdio io.github.ai-mcpower-mcpower-proxy uvx mcpower-proxy