GPT 5.2 does not follow simple instructions?

What version of Codex is running?

codex-cli 0.73.0

What subscription do you have?

pro

Which model were you using?

gpt 5.2 xhigh

What platform is your computer?

Linux

What issue are you seeing?

Why does the AI start exploring random things when asked a simple question?
I literally asked it for a status update since it stupidly reverted uncommitted changes,
"why did you revert the changes in x and y files?
AI : 9 minutes later "exploring" this exploring that, what is the point of exploring hundreds of files when it was asked a very simple question?
literally keeps happening even after I cancelled it, and asked what it was doing it still STARTED EXPLORING FILES!
Explored
└ List tools
Search orb-daily-metrics\.json in scripts
Search orb-symbol-recommendations.json* in data
Read .gitignore, config.ts, time-utils.ts
Search StartOfDay|EndOfDay|ORB start-of-day equity recorded|ORB end-of-day equity recorded in logs

• Ran python3 - <<'PY'
│ import json
│ from pathlib import Path
│ … +10 lines
└ 2025-12-05 ... start 2025-12-05T14:44:20.282Z 8 end 2025-12-05T20:55:08.665Z 962765.57 chg 9991
0.595
… +12 lines
2025-12-16 5dda start 2025-12-16T14:33:04.098Z 952705.97 end None None chg None None
2025-12-16 69 start 2025-12-16T14:45:25.857Z 7175.34 end None None chg None None

• Explored
└ Read risk-manager.ts
Read selector.ts
Search enrichWithSentiment in selector.ts
Read selector.ts
Ran python3 - <<'PY'

What steps can reproduce the bug?

Uploaded thread: 019b232a-62fd-7620-8548-ab91abe84089

What is the expected behavior?

Answer a simple question

Additional information

_No response_