Shared Memory Between ChatGPT and Codex

What feature would you like to see?

Description:
Currently, ChatGPT and Codex run as separate agents without a way to share memory or context across sessions. This limits workflows where a user might start problem-solving in ChatGPT (e.g., brainstorming requirements, storing design notes) and then continue in Codex (e.g., writing and debugging code).

Benefits:
Seamless handoff between natural-language problem framing (ChatGPT) and code implementation (Codex).
Reduced repetition: users don’t have to re-state goals or conventions across agents.
More coherent long-running projects, especially in collaborative or multi-tool workflows.

Are you interested in implementing this feature?

_No response_

Additional information

_No response_