AI Coding Assistants Vulnerable to Malicious Repo Hijacking

By PulseAugur Editorial · [1 sources] · 2026-07-08 16:08

Wiz has identified a critical security vulnerability, dubbed GhostApproval, affecting six AI coding assistants. This flaw allows a malicious repository to hijack the human-in-the-loop confirmation dialog, enabling unauthorized writing to sensitive files such as \"~/.ssh/authorized_keys\". While AWS, Cursor, and Google have already released patches for their respective assistants, Anthropic, Augment, and Windsurf have yet to address the vulnerability. AI

IMPACT This vulnerability could allow malicious code injection into developer workflows, potentially compromising sensitive credentials and systems.

RANK_REASON Security vulnerability disclosure affecting AI developer tools.

Read on Mastodon — fosstodon.org →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

AI Coding Assistants Vulnerable to Malicious Repo Hijacking

COVERAGE [1]

Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] · 2026-07-08 16:08

Wiz disclosed GhostApproval: a symlink flaw in 6 AI coding assistants. A malicious repo can hijack the human-in-the-loop confirmation dialog to write to ~/.ssh/

Wiz disclosed GhostApproval: a symlink flaw in 6 AI coding assistants. A malicious repo can hijack the human-in-the-loop confirmation dialog to write to ~/.ssh/authorized_keys. AWS, Cursor, and Google patched; Anthropic, Augment, Windsurf have not. https:// go.aintelligencehub.co…

LINKS aintelligencehub.com/link-not-found aintelligencehub.com/link-not-found

COVERAGE [1]

Wiz disclosed GhostApproval: a symlink flaw in 6 AI coding assistants. A malicious repo can hijack the human-in-the-loop confirmation dialog to write to ~/.ssh/

RELATED ENTITIES

RELATED TOPICS