Sandbox & Runtimedsh-plugin

hol-guard

HOL Guard is an open-source antivirus for AI agents that acts as a local-first security layer for agents, tools, plugins, skills, MCP servers, and package installs. It reviews actions before they run, detects secret exposure, destructive operations, prompt injection, and supply-chain risks, and allows, blocks, or requests approval based on policy.

Stars
738
Forks
132
Open issues
137
Last push
Oct 3, 2026
Latest release
v3.18.1
24h Growth
+26
7d Growth
+70

Installation

No reliable install command was detected. Check the project README for installation instructions.

Overview

HOL Guard is an open-source antivirus for AI agents that acts as a local-first security layer for agents, tools, plugins, skills, MCP servers, and package installs. It reviews actions before they run, detects secret exposure, destructive operations, prompt injection, and supply-chain risks, and allows, blocks, or requests approval based on policy.

  • Local-first security layer for AI agents, tools, plugins, skills, MCP servers, and package installs.
  • Reviews shell commands, file access, package installs, and MCP tool calls before they run.
  • Detects secret exposure, destructive operations, prompt injection, and supply-chain risks.
  • Allows, blocks, or requests approval according to policy.
  • CLI and local dashboard manage protection, resolve approvals, and inspect decision history.
  • Optional Guard Cloud adds shared history, team policy, and fleet management.
  • Plugin Scanner CLI runs security and quality checks for agent plugins, skills, and MCP integrations.
  • Scans Codex, Claude Code, DeepSeek Harness, Gemini CLI, Kimi Code, and OpenCode package formats.

Related plugins