Plugin4Shell Flaw Hits Four AI Coding Agents — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Air Security disclosed on Thursday a flaw it calls Plugin4Shell affecting four widely used AI coding agents, in which anyone who controls a plugin's code repository can swap a plugin an agent installs for a malicious one even after the agent pinned that plugin to a specific reviewed version.
- Anthropic is named among the AI-coding-agent vendors Air Security briefed in its Thursday disclosure (the source excerpt is truncated mid-sentence while describing Anthropic's response).
Why it matters: Version-pinning is the basic trust control developers rely on when pulling third-party plug-ins into AI coding assistants; Air Security's finding means a single plugin maintainer can silently turn a pinned, reviewed build into a malicious one at install time across four major agent platforms, shifting the threat model from malicious publishers to compromised or coerced legit ones.
Ask SkimNews
