Want to try OpenClaw? NanoClaw is a simpler, potentially safer AI agent

NanoClaw is an open-source, lightweight AI agent positioned as a more secure alternative to OpenClaw: it uses a much smaller, auditable codebase and runs each agent in isolated containers by default. The report highlights security advantages such as reduced attack surface, Claude-based mitigation against prompt injection, and recommends strict admin-group controls and disabling internet/search for the main agent, while warning that residual risks remain and caution is still necessary.