Shell

关于

Shell offers a Model Context Protocol (MCP) server that enables secure execution of shell commands. It features advanced command whitelisting, optional user confirmation, and robust security controls, making it ideal for developers and AI agents requiring a safe and flexible way to interact with the shell across various transports like stdio, SSE, and streamable HTTP. It also prevents hanging processes with timeout protection and provides context-aware error messages for better agent guidance.

主要功能

• Advanced command whitelisting with flexible allow/deny policies
• User confirmation for non-whitelisted commands via MCP elicitation
• Support for multiple transports (stdio, SSE, streamable HTTP)
• Timeout protection prevents hanging processes
• Shell feature restrictions (pipes, redirects, substitutions, etc.)
• 0 GitHub stars

使用案例

• Providing a controlled environment for developers to test and run shell scripts
• Securing automated workflows that require shell command execution
• Enabling LLM agents to safely execute shell commands under strict policy controls