Critical Codex data-loss incident — compensation requested

What version of Codex CLI is running?

codex 0.149.1

What subscription do you have?

Pro x20

Which model were you using?

sol ultra

What platform is your computer?

Windows 11 x64

What terminal emulator and version are you using (if applicable)?

No response

Codex doctor report

What issue are you seeing?

Critical Codex data-loss incident — compensation requested

On August 23, 2026, during a Codex Heavy-route goal, a Codex subagent
executed a malformed Windows rmdir command intended only for a
pycache directory. Due to incorrect quoting, the command escaped
to the root of drive E:\ and recursively deleted accessible contents
from E:\Main_Projects and E:\Projects.

The command ran approximately from 23:07:43 to 23:18:44 Europe/
Bucharest. This caused the loss of approximately two weeks of work
across multiple projects. The files bypassed the Recycle Bin.

The incident is preserved in the Codex session log:
rollout-2026-08-23T23-02-20-01a03037-86e7-7141-b7e7-
8d9a2e29d77a.jsonl

Session evidence SHA-256:
21e18da945a41b12a55f17a90816f81a6099f599d60cc6738097f6c4212f426d

I request:

1. Immediate escalation to the Codex safety/engineering team.
2. A full reset of my banked Codex usage.
3. Account credits or an appropriate refund/compensation.
4. A written incident/case number.
5. Confirmation that the malformed destructive-command behavior will
   be investigated.

What steps can reproduce the bug?

Uploaded thread: 01a031f9-a65f-7141-82c5-c27b20d06c32

What is the expected behavior?

No response

Additional information

No response

Start by reviewing the uploaded thread and the named Codex session log, rollout-2026-08-23T23-02-20-01a03037-86e7-7141-b7e7-8d9a2e29d77a.jsonl. No repository file or test is identified, and the expected behavior is not stated; done would require a confirmed engineering investigation of the malformed destructive command and its safety implications.