Welcome! Register below to get the join link and a calendar invite by email.
There's a new "Claw" tool every week, and it's hard to know which ones are worth your time, or which ones are safe to let near your files and accounts. NanoClaw is one of the simplest: a free, open-source AI agent that runs on your own computer, and you talk to it from apps you already use, like Slack, WhatsApp or Telegram.
In this session I'm joined by Avital Weisinger from the NanoClaw team. Together we'll look at what can go wrong when you give an AI agent access to your work, show how NanoClaw is built to prevent it, and set one up live in Slack.
What we will cover
- What NanoClaw is (and isn't): a personal agent you run yourself, and how it differs from the other Claw tools
- The security risks of AI agents: an agent that can read your files, run code and use your accounts can also leak data, follow hidden instructions planted in an email or web page, or do something you never meant to approve
- How NanoClaw keeps you in control: each agent runs in its own sealed box and sees only the files you give it, your passwords and keys never reach the agent, sensitive actions wait for your approval, and the code is small enough to check
- Live setup in Slack: from a fresh install to your first conversation with your agent
- Real uses: a morning briefing, a weekly review, and separate agents for work and personal life
- Your questions: open Q&A with Avital
Who it is for
Anyone who uses ChatGPT or Claude as a chat and wants an assistant that works for them in the background, without handing it the keys to everything. Setup runs in a terminal; we'll go through it step by step.
Last updated: