Weekly Build: OpenClaw
68,000 GitHub stars. 700 million users through Baidu. 135,000 agents accidentally exposed to the internet. OpenClaw is having a moment.
OpenClaw is the open-source personal AI agent that runs locally, connects to your existing messaging apps (WhatsApp, Slack, Telegram, iMessage), and can actually do things on your machine — manage files, execute commands, automate workflows. No subscription, bring your own API key. It might be the first AI agent that regular people are actually using day to day.
But it's also raising real questions. Researchers just found over 135K agents exposed online with remote code execution risks, and there are concerns about malicious skills and prompt injection.
This Friday at Weekly Build, we're talking about OpenClaw — what people are actually using it for, whether it lives up to the hype, and what the security story means for anyone thinking about running it. If anyone in the group has been using it, we'd love to hear your experience.
What to expect:
What OpenClaw actually is and why it's blowing up right now
Real talk about what's useful vs. what's hype
The security side — what to know before you run an AI agent on your machine
How it fits into the bigger picture alongside Claude Code, Codex, and other AI tools
Come hang whether you've been running OpenClaw for weeks or just heard about it for the first time reading this. No expertise required — we're all figuring this out together.
Hosted by ShipSLC. We build with the latest tools every Friday and keep each other shipping.