Potential Security Risk: AI Model Executing Shell Commands Without Strict Validation

Resolved 💬 2 comments Opened Feb 28, 2025 by changjonathanc Closed Mar 13, 2025

Bug Description
claude decides to run cd .... && python ..... and the system ask me to approve, or approve all cd commands, but i think this is an issue if it allow model to run any commands after cd

Environment Info

  • Platform: macos
  • Terminal: kitty
  • Version: 0.2.27
  • Feedback ID: 2e373c06-c41e-4a63-b7ac-80abdf99b1e6

View original on GitHub ↗

This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗