ShellGuard FAQs

Question 1

What kind of tasks can LLM agents perform using ShellGuard?

Accepted Answer

LLM agents can use ShellGuard to connect to servers, execute read-only commands for diagnostics, inspect logs, download files (with limits), and manage SSH connections, mimicking a human's manual troubleshooting workflow.

Question 2

How does ShellGuard ensure security for remote server access?

Accepted Answer

ShellGuard enforces a robust command validation pipeline, allowing only a curated set of read-only operations. Destructive commands are blocked with actionable suggestions, enabling LLMs to self-correct and maintain system integrity.

Question 3

Which SSH features and clients does ShellGuard support?

Accepted Answer

ShellGuard supports comprehensive `~/.ssh/config` features, including `ProxyJump` and `Match` blocks, via both native Go SSH and system OpenSSH modes. It integrates seamlessly with MCP clients like Cursor, Claude, OpenCode, VS Code, Zed, and Roo Code.

Question 4

Can ShellGuard provision diagnostic tools on remote servers?

Accepted Answer

Yes, ShellGuard enables on-demand provisioning of essential diagnostic tools like `ripgrep`, `jq`, and `yq` on remote hosts, ensuring LLMs have the necessary utilities for effective troubleshooting.

Question 5

What is ShellGuard and how does it help LLM agents?

Accepted Answer

ShellGuard is an MCP server that empowers LLM (Large Language Model) agents with secure, read-only bash access to remote servers over SSH. This allows AI to perform diagnostics, inspect logs, query databases, and troubleshoot hands-free.

ShellGuard

ShellGuard

Key Features

Use Cases

Key Features

Use Cases