Severity: CriticalVulnerabilityModel/inference
Multiple AI coding tools vulnerable to sandbox escape via untrusted file execution
Global
Live intelligence. Items are aggregated from public sources and summarised automatically. Always verify against the linked source before acting.
Researchers discovered sandbox escape vulnerabilities in several AI-assisted coding platforms—Cursor, Codex, Gemini CLI, and Antigravity—by exploiting the ability of AI agents to write files that host tools subsequently execute without validation. The attack chain leverages the trust relationship between the AI sandbox and the host environment to achieve code execution outside intended boundaries.
What to do
Isolate AI agent file output from host tool execution paths and validate all files before trusted execution.
Mapped framework pillars
Sources
#sandbox escape#AI agents#code execution#file write#trusted execution