117
Views
0
Saves
About HOL Guard
HOL Guard is an open-source runtime security layer for AI coding agents. It reviews shell commands, file access, package installs, and MCP tool calls before they run, allowing, blocking, or prompting for approval based on your policy. The local runtime is free and open-source under Apache-2.0; optional Guard Cloud plans add sync and team features.
Updated 10/30/2026
Key Features
- Local-first runtime protection for AI agents
- Blocks or prompts before destructive shell commands, file access, and secret exposure
- Scans package installs for supply-chain risks
- Inspects plugins, skills, MCP servers, and agent configs
- Prompt-injection detection in tool results
- Free and open-source (Apache-2.0) — no account required for local use
- Supports Codex, Claude Code, Cursor, OpenCode, Hermes, and more
- Optional Guard Cloud for sync, alerts, and team policy
Pricing Structure
- Free — $0 — Local protection forever; CLI, dashboard, approvals, offline support
- Solo — $4.99/month — 2 cloud devices, 30-day history, 1 GB storage, weekly digest
- Pro — $15/month — 5 devices, 180-day history, 5 GB storage, real-time alerts, policy versioning
- Team — $30/seat/month — Up to 25 devices, 365-day history, 25 GB + 5 GB/seat, shared policy
Why People Use It
HOL Guard is popular because it puts a local security checkpoint between AI coding agents and the machine, catching risky actions before they run instead of auditing them after the damage is done. The open-source core means teams can inspect and extend the policy, while the optional cloud tier adds the fleet-management features enterprises actually need.
See HOL Guard in Action
0
selected