GitLost: We Tricked GitHub's AI Agent into Leaking Private Repos
Noma Labs discovered a critical prompt injection vulnerability, dubbed 'GitLost,' in GitHub's new Agentic Workflows, allowing an unauthenticated attacker to silently extract data from private repositories. This is achieved by posting a specially crafted GitHub Issue in a public repository belonging to the same organization as the private ones.
Why it matters
This vulnerability highlights significant security risks associated with integrating AI agents into critical development infrastructure, demonstrating how prompt injection attacks can bypass intended security measures and lead to sensitive data breaches, impacting data privacy and corporate security.
TL;DR : Noma Labs discovered a critical prompt injection vulnerability within GitHub’s new Agentic Workflows, allowing an unauthenticated attacker to silently pull data from private repositories by posting a crafted GitHub Issue in a public repository belonging to the same organization as the private repositories. Noma Labs named the vulnerability GitLost.
Hacker News is a tech news aggregator and community. The article is a technical report on a security vulnerability, presented factually without political or ideological bias.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in