New Attacks Trick OpenClaw AI Agent Into Running Code and Leaking Secrets
OpenClaw input flaws let hidden contacts and phishing emails trigger code execution and data leaks, exposing agent trust risks.
More info