Guides

On-device agent security · practical guides

Is OpenClaw safe? 7 things to check before self-hosting

Self-hosted agents like OpenClaw put your keys and execution rights on your own machine — unlocked, that’s a naked digital butler. These 7 are the security floor before you deploy.

Read →

The 5 biggest security risks of on-device AI agents (and how to test)

Self-hosted / on-device agents face five risk classes: exposed admin panel, supply-chain poisoning, prompt injection, confused deputy, and credential blast radius. Explained, with how to test each.

Read →