DefenseClaw

DefenseClaw

Security governance for agentic AI

Description

Letting agents like OpenClaw call tools freely means nobody knows what they access or send, and there's no audit trail. DefenseClaw from Cisco is security governance for OpenClaw and agentic AI runtimes.

It scans capabilities before use, inspects runtime traffic and exports durable audit evidence, combining a Python CLI and Go gateway.

Features



Capability scans:Before use.

Traffic inspection:At runtime.

Audit evidence:Durable.

Gateway:Central control.