"Shell" matching MCP servers:

An agent-native teaching OS that gives AI assistants a persistent classroom with courses, lessons, flashcards, and graded exercises via MCP tools.

Enables executing local commands via an argv-based tool with optional non-interactive sudo, command allowlisting, and filesystem write confinement using Bubblewrap.

Enables AI agents to monitor server health and capacity, diagnose outages, inspect Docker and deployment status, and perform safe, bounded recovery actions via MCP without granting unrestricted shell or SSH access.

More than just an MCP server that covers the entire range of Docker management tasks, it helps you create workflows to easily manage your Docker environments at scale. It allows you manage multiple local or remote (TCP, TLS, SSH) docker instances and exposes things like logs and stats as resources so that you can monitor and triage all of them.

MCP server that vets LLM-emitted shell commands BEFORE execution — detects rm -rf nested deep in chains, package-manager glob removal (apt remove 'nvidia'), dd/mkfs filesystem destruction, chmod 777 / chown -R privilege blast, network-exfil via curl | bash, chained shutdown/reboot, git destructive ops. 30 detection rules across 8 families. Sub-second, local, free, MCP-native.

Your sysadmin co-pilot — an AI that administers Linux through typed, approval-gated, Ed25519-audited actions instead of shell strings. Reference implementation of the LACS standard.

Outbound-only remote shell, detached long-running jobs, and temporary file courier for AI agents. Hosted Streamable HTTP plus stdio via npx -y @aicommander/mcp.

Indexes git history, shell commands with exit codes, project docs, and opt-in assistant transcripts into a local SQLite database, then serves ranked, token-budgeted context back over MCP. Uniquely links a failed shell command to the commit that fixed it, so agents can recall what was already tried and what worked.

Standard coding agents waste 70% of context tokens re-reading raw files with shell grep. Synapse-MCP indexes your codebase into a sub-millisecond local AST graph—giving Claude Code, Cursor, and Windsurf exact caller trees, 60% lower token spend, and zero-hallucination refactoring.

Advanced shell command execution for AI agents — a Model Context Protocol (MCP) server with security policy engine, process lifecycle management, bounded output capture, POSIX signals, and token-efficient responses.

A Model Context Protocol server that allows LLMs to securely execute shell commands on remote Linux and Windows systems via SSH. It supports password and key-based authentication, command timeouts, and sudo elevation for administrative tasks.

Secure, read-only MCP server that lets AI assistants (Claude, OpenCode, VS Code) inspect Android devices via ADB — logcat, UI hierarchy, installed packages, and allowlisted shell commands. TypeScript/Node.js.

Enables sub-millisecond pre-execution safety filtering for AI agent tool calls, blocking destructive shell commands, dangerous SQL mutations, credential access, SSRF, and scope-creep or prompt-injection patterns before they execute, with optional fail-closed transparent proxy wrapping for any MCP server.

Enables AI models to dynamically create and execute their own custom tools through a meta-function architecture, supporting JavaScript, Python, and Shell runtimes with sandboxed security and human approval flows.

A local stdio MCP server wrapping arduino-cli to let agents detect boards, manage cores/libraries, compile and upload sketches, and communicate over serial, enabling hardware control without shell gymnastics.

Provides LLMs with safe, controllable, and stateful SSH access to shell environments, enabling remote command execution and file transfers with async support and security features.

Bareun (바른) is a Korean natural-language engine. This MCP server lets any MCP-compatible client hand Korean text to a dedicated Korean engine instead of guessing: analyze_syntax (morphological analysis with 47 POS tags and beta homograph sense disambiguation), analyze_syntax_raw (raw model output), tokenize, correct_grammar (spelling and spacing correction), and list_pos_tags. Hosted at https://ap

Local macOS MCP bridge that gives ChatGPT shell, unrestricted filesystem access, real PTY sessions, background jobs, and read-only stored Codex history. ChatGPT stays the reasoning layer; the bridge makes no model calls.

Installs an agent on Linux hosts to allow shell commands, file edits, and service management through Claude.ai or ChatGPT via a secure outbound WebSocket connection.

MCP server that enforces deterministic policies for coding agents, inspecting file writes and shell commands to allow, deny, or require confirmation based on configurable rules.