Professor Shares Prompt Trick to Stop OpenAI Codex from Delegating

Professor Ethan Mollick shared a prompt trick for OpenAI Codex after finding it frequently delegates tasks to less capable sub-agents. Users can issue forceful commands to prevent this delegation and improve efficiency.

2026-08-09 ~ 2026-08-09 · 2 related posts

• Stop Delegating to Dumber Agents: Real-World Codex Pain Point — emollick · 2026-08-09
• Professor Tests Codex: One Prompt to Stop AI from Delegating to Dumber Agents — emollick · 2026-08-09