-mcp
by syphrpunk
Overview
A Model Context Protocol server for executing any Open Source tool via `pkgx` within a sandboxed environment on macOS, or unsandboxed on Linux.
Installation
npx ts-node servers/pkgx/main.tsSecurity Notes
The server explicitly states it is a 'dangerous tool' that allows AI to read almost any file on the user's computer. While it implements macOS-specific sandboxing to prevent arbitrary file writes (except to /tmp, /var, /private), it does not restrict file reads (except ~/.ssh, ~/.aws). Crucially, there is no sandboxing implemented for Linux, making it highly vulnerable to arbitrary command execution with full file system access on that platform. Execution of arbitrary commands (via `execSync` and `spawn`) based on AI input is a severe risk, despite attempts to parse command lines and prevent root execution.
Similar Servers
XcodeBuildMCP
Provides an MCP (Model Context Protocol) server for AI agents and other clients to programmatically build, test, run, and debug iOS and macOS applications, manage simulators/devices, and capture logs.
hyper-mcp
A fast, secure Model Context Protocol (MCP) server that extends its capabilities through WebAssembly plugins, enabling AI agents to access tools, resources, and prompts.
mcp-server-code-execution-mode
This server enables LLM agents to execute Python code in a highly secure, isolated container environment, facilitating complex multi-tool orchestration and data analysis with minimal LLM context token usage.
orla
Orla acts as a runtime for Model Context Protocol (MCP) servers, enabling the execution of lightweight open-source AI agents and command-line tools locally.