AI coding assistants leak CI secrets through GitHub issue flaws

THE article discusses vulnerabilities found in Claude Code and Gemini CLI that allowed GitHub issues to access Continuous Integration (CI) workflow secrets. It emphasizes the risks associated with AI-generated code and highlights the need for tighter security measures during software development. The piece points out that these vulnerabilities could facilitate potential data breaches, urging developers to adopt better governance and risk management strategies in the face of rapid AI advancements.