# HOL Guard

> Apache local antivirus for agents — MCP, skills, shell, packages

**Category:** AI Agents  
**Pricing:** open-source  
**URL:** https://hol.org/guard  
**GitHub:** https://github.com/hashgraph-online/hol-guard  
**Added:** 2026-10-02

---

## The Hook

HOL Guard is an Apache local-first security layer for AI agents — review shell, files, package installs, and MCP tool calls before they run, with allow/block/approve policy.

## What's Good

Apache-2.0 (hashgraph-online/hol-guard). PyPI `hol-guard` and `plugin-scanner` (HTTP 200). `pipx install hol-guard` then `hol-guard init` — no account required for local mode. Covers secret exposure, destructive ops, prompt injection, and supply-chain risks across agents/plugins/skills/MCP. Optional Guard Cloud for team policy is separate from the OSS bar.

## The Catch

Vs SkillSpector / mcp-scanner / destructive-command-guard / AI-Infra-Guard: SkillSpector scans skills pre-install; MCP Scanner audits MCP servers; dcg is a host deny-hook; AI-Infra-Guard is red-team platform — HOL Guard’s catch is runtime local policy/antivirus across agent actions including MCP calls. Policy false positives can stall agents; optional cloud is not required — keep local-first if that is the point.

## Verdict

Apache local agent antivirus when you need runtime allow/block on tools and MCP — not skills-only scanners and not MCP-server audit alone.