Agent Panel: Agent doesn't respect diagnostics in Rust.

[poly2it]

Summary

Agent doesn't respect diagnostics in Rust.

Description

Steps to trigger the problem:

1. On a minimal Zed install, with GPT-4.1 via Github Copilot with diagnostics enabled
2. Asking for edits in a Rust project
3. The model does not attempt to repair code when diagnostics show clear errors

Actual Behavior: The model doesn't receive and diagnostics. Expected Behavior: The model should receive continious diagnostics to hinder it from messing up.

Zed Version and System Specs

Zed: v0.186.5 (Zed Preview) OS: Linux Wayland nixos 25.05 Memory: 62.6 GiB Architecture: x86_64 GPU: NVIDIA GeForce RTX 3060 || NVIDIA || 570.133.07

[agu-z]

@poly2it Does the model not run the diagnostics tool at the end?

[osyvokon]

Currently, we expect the model to ask for diagnostics. Alternatively, we may start sending newly introduced diagnostics to the model proactively. Need to validate that this approach doesn't introduce too much noise first (especially for languages like Python, where diagnostics are often noisy)

[poly2it]

@poly2it Does the model not run the diagnostics tool at the end?

Yes. Seemingly from the adverts, it did to ensure that it would not create broken code.

[github-actions[bot]]

Hi there! 👋 We're working to clean up our issue tracker by closing older bugs that might not be relevant anymore. If you are able to reproduce this issue in the latest version of Zed, please let us know by commenting on this issue, and it will be kept open. If you can't reproduce it, feel free to close the issue yourself. Otherwise, it will close automatically in 14 days. Thanks for your help!

[poly2it]

.