Hol Guard
Open-source antivirus for AI agents: block risky tools, secret access, prompt injection, malicious packages, MCP servers
Open-source antivirus for AI agents: block risky tools, secret access, prompt injection, malicious packages, MCP servers, plugins, and skills at runtime.
Setup and access
Check the connection method, access level, and compatible clients before enabling the server.
- Transport
- stdio
- Authentication
- Varies
- Registry status
- Active
- Last updated
- Aug 8, 2026
Connects to external tools and data. Use the least-privileged credentials and access scope that support your workflow.
FAQ
- What does Hol Guard MCP do?
- Open-source antivirus for AI agents: block risky tools, secret access, prompt injection, malicious packages, MCP servers, plugins, and skills at runtime.
- How do I connect Hol Guard MCP?
- Follow the stdio setup in the source repo, then confirm permissions before you enable it.
- Does Hol Guard MCP require authentication?
- It depends on the capability you enable. Check the setup notes for the exact service path.
- Who maintains Hol Guard MCP?
- hashgraph online publishes this implementation. Check the source repo for maintainers and releases.
- Where is the Hol Guard source?
- Source and updates: https://github.com/hashgraph-online/hol-guard