A critical vulnerability has been discovered in GitHub's Agentic workflows, allowing attackers to inject malicious prompts and expose sensitive data from private repositories without requiring authentication. This is made possible by crafting a public GitHub Issue that tricks the AI-powered workflow into divulging confidential information. The attack exploits a weakness in the workflow's design, enabling unauthorized access to private data. Specifically, an attacker can create a carefully designed Issue that, when processed by the workflow, reveals sensitive information from private repositories. This vulnerability has significant implications for organizations relying on GitHub for development and collaboration, as it could lead to unauthorized data exposure1. The discovery of this vulnerability highlights the importance of securing AI-powered workflows and ensuring the integrity of private repositories. So what matters most to practitioners is that they must prioritize the security of their workflows to prevent such attacks and protect sensitive data.