Monday, March 2nd
11:00 AM PT / 2:00 PM ET
AI agents are moving fast. Attackers are moving faster.
OpenClaw is powerful. It can also hand LLMs deep control over your environment. In many deployments, that effectively means giving an AI agent root level influence on a machine. What could go wrong? Join Jimmy Mesta, CTO and Co-Founder of RAD Security, and Ben Sadeghipour, CEO and Co-Founder of HackingHub, for a live demo and public launch of Clawkeeper, the first security scanner and hardening platform built specifically for OpenClaw AI agent deployments.
The Problem
In real OpenClaw environments, we keep finding the same attack paths:
- Plaintext API keys and tokens sitting in config files
- Unvetted marketplace skills running with full agent privileges
- Remote code execution paths that allow arbitrary code on the host
- Config drift that quietly increases blast radius across cloud accounts
From a red team perspective, these are not theoretical risks. They are direct paths from agent misconfiguration to host level impact.
AI agents are now a new attack surface. Most security teams already juggle cloud, identity, and CI pipelines. In many organizations, no one fully owns AI agent security yet, and that gap is widening.
The Solution: Clawkeeper
Clawkeeper scans from host to agent in under 60 seconds. Beyond just flagging issues, it shows you how an attacker would actually chain them together. You get:
- A clear security grade
- A prioritized list of what to fix first
- Visibility into real exploit paths
- Hardening guidance tailored specifically to OpenClaw
In this session, we will:
- Break down how OpenClaw works under the hood and where the sharp edges are
