[Bug] Model exhibits overconfident reasoning without knowledge verification capability

Bug Description
I think the llm stays with the current knowledge and even though it might be wrong, it continues to claim it is right. No hint of curiosity. Fortunately, I know where to look to validate if new data is available.

Environment Info

• Platform: linux
• Terminal: gnome-terminal
• Version: 2.1.80
• Feedback ID: ee9e9a5e-812e-4472-a2cb-591ea818cb6e

Errors

[{"error":"Error: Request was aborted.\n    at makeRequest (/$bunfs/root/src/entrypoints/cli.js:255:3940)\n    at processTicksAndRejections (native:7:39)","timestamp":"2026-03-20T21:03:26.174Z"}]

It was going to implement a very complex function to extract the dissolution gas for a co2 storage simulation because the keywords have been recently added. But if they provide a functionality to calculate something, they might have provided a function to read it as well. So it would be nice to be able to update his knowledge before trying to implement some heavy function to solve the issue.
just saying...Thank...great tool!