Fine grained context per task instead of per project.
Resolved 💬 4 comments Opened Mar 4, 2026 by tradingfuturo Closed Apr 18, 2026
💡 Likely answer: A maintainer (github-actions[bot], contributor)
responded on this thread — see the highlighted reply below.
What variant of Codex are you using?
IDE Extension, CLI
What feature would you like to see?
Currently all AI agents are architected on a per project basis. This causes them to make exploration of the project a large and time-consuming task. If users could feed agents with more specific per task context a lot of time and tokens would be saved per task. Example, have several setups per project with agents.md + mcp + files + prompt (or mcp + files + prompt) for each specific task. I could select one of those prepared contexts (example with a dropdown), add the prompt and run the task. The result on average less exploration, less tokens and less time = faster development speed.
Additional information
https://github.com/anthropics/claude-code/issues/20292#event-23168262041
4 Comments
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
You can already do this using skills that you invoke or simple ".md" files that you @-mention in your prompt. Have you tried those approaches?
I have not tried skills in codex. But as for the other yes something similar, is the next best thing. I use slim agents.md with paths to other .md files so AI saves exploration time and automatically reads those md when required during exploration (example architecture.md or testing.md). But seems there is no quick mcp reconfiguration. Not sure if this can be achieved with skills.
This feature request hasn't received any upvotes in over a month, so I'm going to close it.