mirror of
https://github.com/openai/codex.git
synced 2026-02-01 22:47:52 +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