OpenClaw surpasses Linux and React as GitHub's most-starred project. Security experts warn its agentic AI architecture poses enterprise risks without controls.

The New Stack is a publication covering trends and technologies in cloud-native development, DevOps, and software delivery. Developers can learn about containerization, Kubernetes, and cloud computing, as well as explore topics such as microservices architecture, serverless computing, and continuous integration/continuous delivery (CI/CD) pipelines.

The New Stack

OpenClaw, an open-source autonomous AI agent, has surpassed Linux and React to become GitHub's most-starred non-aggregator software project, accumulating 250K+ stars in just four months. The tool acts as a personal AI assistant running locally, integrating with messaging platforms like WhatsApp, Slack, and Teams. Despite its explosive popularity, security experts warn that its agentic architecture introduces serious enterprise risks including prompt injection, unsafe tool invocation, and lack of fine-grained access controls. Its third-party skills marketplace ClawHub was already hacked twice in early 2026. Experts call for embedding control planes, centralized authentication, runtime guardrails, and observability into agentic traffic before enterprise adoption scales further.

OpenClaw rocks to GitHub’s most-starred status, but is it safe?