mirror of
https://github.com/openai/codex.git
synced 2026-04-27 08:05:51 +00:00
helpful in the future if we want more granularity for requesting escalated permissions: e.g when running in readonly sandbox, model can request to escalate to a sandbox that allows writes