Have the agent be able to see/read terminal output

Feature Requests
1

Please give the agent the ability ot see the terminal output, and allow us to @ reference terminal output (though even better let us provide custom plugins/context providers).

Most of the competing coding agent frameworks give the agent full access to terminal outputs

1 Like
2

Adding a vote for this. My specific use case (for now) is as follows:

  1. I am developing a react native application, and running the dev server in a terminal window within cursor ( npm run ios )
  2. Im using the cursor agent to make most code changes
  3. I want the agent to monitor/automatically read the logs in the terminal while it is making a change so it can debug itself.

I envision this working similar to how linting within the agent works.

3

Hey, in v0.46, you can now @ a terminal directly.

I don’t believe the Agent can choose to read them yet, mainly because you could have 5 terminals and non of them might be running anything, so it would be pointless to check them!