edit_prediction: Skip cloud Zeta requests when not signed in#57615
Merged
Conversation
When the user is not signed in, the edit prediction system was attempting a cloud API request on every keystroke, failing deep in the credential check with a noisy ERROR log. Return early before building inputs or spawning any tasks.
The previous guard used current_user() which is None in tests even when credentials are set directly on the cloud client. has_credentials() is the accurate check for whether a cloud request can proceed.
5 tasks
Co-authored-by: David3u <3udavid@gmail.com>
Member
|
/cherry-pick stable |
zed-zippy Bot
added a commit
that referenced
this pull request
May 29, 2026
… (cherry-pick to stable) (#58044) Cherry-pick of #57615 to stable ---- Self-Review Checklist: - [x] I've reviewed my own diff for quality, security, and reliability - [x] Unsafe blocks (if any) have justifying comments - [x] The content is consistent with the [UI/UX checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist) - [x] Tests cover the new/changed behavior - [x] Performance impact has been considered and is acceptable Closes #57962 ## What When a user is not signed in to their Zed account, the edit prediction system was still attempting a cloud API request on every keystroke. The request would fail deep in the credential check (`CloudApiClient::build_request`) with a `ClientApiError::NotSignedIn` error, which propagated back up and was logged at `ERROR` level via `.log_err()` at line 2389 of `edit_prediction.rs`. ## Why The sign-in check was happening too late — only discovered after async tasks were already spawned and the full request pipeline entered. This fix gates the request at the top of `request_prediction_internal`, returning `Task::ready(Ok(None))` immediately before any inputs are built or tasks spawned. The guard mirrors the existing `is_cloud` provider check already used elsewhere in the same file, and only applies to the `Zeta` model on the cloud provider path. Local providers (Ollama, `OpenAiCompatibleApi`) and other models (Mercury, Fim) are unaffected. Note: I haven't added a test for this — testing the early-return would require mocking auth state, which I wasn't sure was worth the complexity for a one-liner guard. Happy to add one if preferred. Release Notes: - Fixed noisy `not signed in` error log on every keystroke when not signed in to Zed --------- Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev> Co-authored-by: MrSubidubi <finn@zed.dev> Co-authored-by: David3u <3udavid@gmail.com> Co-authored-by: Seth Wood <sethlwood@me.com> Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev> Co-authored-by: MrSubidubi <finn@zed.dev> Co-authored-by: David3u <3udavid@gmail.com>
zed-zippy Bot
added a commit
that referenced
this pull request
May 29, 2026
… (cherry-pick to preview) (#58045) Cherry-pick of #57615 to preview ---- Self-Review Checklist: - [x] I've reviewed my own diff for quality, security, and reliability - [x] Unsafe blocks (if any) have justifying comments - [x] The content is consistent with the [UI/UX checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist) - [x] Tests cover the new/changed behavior - [x] Performance impact has been considered and is acceptable Closes #57962 ## What When a user is not signed in to their Zed account, the edit prediction system was still attempting a cloud API request on every keystroke. The request would fail deep in the credential check (`CloudApiClient::build_request`) with a `ClientApiError::NotSignedIn` error, which propagated back up and was logged at `ERROR` level via `.log_err()` at line 2389 of `edit_prediction.rs`. ## Why The sign-in check was happening too late — only discovered after async tasks were already spawned and the full request pipeline entered. This fix gates the request at the top of `request_prediction_internal`, returning `Task::ready(Ok(None))` immediately before any inputs are built or tasks spawned. The guard mirrors the existing `is_cloud` provider check already used elsewhere in the same file, and only applies to the `Zeta` model on the cloud provider path. Local providers (Ollama, `OpenAiCompatibleApi`) and other models (Mercury, Fim) are unaffected. Note: I haven't added a test for this — testing the early-return would require mocking auth state, which I wasn't sure was worth the complexity for a one-liner guard. Happy to add one if preferred. Release Notes: - Fixed noisy `not signed in` error log on every keystroke when not signed in to Zed --------- Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev> Co-authored-by: MrSubidubi <finn@zed.dev> Co-authored-by: David3u <3udavid@gmail.com> Co-authored-by: Seth Wood <sethlwood@me.com> Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev> Co-authored-by: MrSubidubi <finn@zed.dev> Co-authored-by: David3u <3udavid@gmail.com>
TomPlanche
pushed a commit
to TomPlanche/zed
that referenced
this pull request
Jun 2, 2026
…ustries#57615) Self-Review Checklist: - [x] I've reviewed my own diff for quality, security, and reliability - [x] Unsafe blocks (if any) have justifying comments - [x] The content is consistent with the [UI/UX checklist](https://github.com/zed-industries/zed/blob/main/CONTRIBUTING.md#uiux-checklist) - [x] Tests cover the new/changed behavior - [x] Performance impact has been considered and is acceptable Closes zed-industries#57962 ## What When a user is not signed in to their Zed account, the edit prediction system was still attempting a cloud API request on every keystroke. The request would fail deep in the credential check (`CloudApiClient::build_request`) with a `ClientApiError::NotSignedIn` error, which propagated back up and was logged at `ERROR` level via `.log_err()` at line 2389 of `edit_prediction.rs`. ## Why The sign-in check was happening too late — only discovered after async tasks were already spawned and the full request pipeline entered. This fix gates the request at the top of `request_prediction_internal`, returning `Task::ready(Ok(None))` immediately before any inputs are built or tasks spawned. The guard mirrors the existing `is_cloud` provider check already used elsewhere in the same file, and only applies to the `Zeta` model on the cloud provider path. Local providers (Ollama, `OpenAiCompatibleApi`) and other models (Mercury, Fim) are unaffected. Note: I haven't added a test for this — testing the early-return would require mocking auth state, which I wasn't sure was worth the complexity for a one-liner guard. Happy to add one if preferred. Release Notes: - Fixed noisy `not signed in` error log on every keystroke when not signed in to Zed --------- Co-authored-by: Oleksiy Syvokon <oleksiy@zed.dev> Co-authored-by: MrSubidubi <finn@zed.dev> Co-authored-by: David3u <3udavid@gmail.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Self-Review Checklist:
Closes #57962
What
When a user is not signed in to their Zed account, the edit prediction system was still attempting a cloud API request on every keystroke. The request would fail deep in the credential check (
CloudApiClient::build_request) with aClientApiError::NotSignedInerror, which propagated back up and was logged atERRORlevel via.log_err()at line 2389 ofedit_prediction.rs.Why
The sign-in check was happening too late — only discovered after async tasks were already spawned and the full request pipeline entered. This fix gates the request at the top of
request_prediction_internal, returningTask::ready(Ok(None))immediately before any inputs are built or tasks spawned.The guard mirrors the existing
is_cloudprovider check already used elsewhere in the same file, and only applies to theZetamodel on the cloud provider path. Local providers (Ollama,OpenAiCompatibleApi) and other models (Mercury, Fim) are unaffected.Note: I haven't added a test for this — testing the early-return would require mocking auth state, which I wasn't sure was worth the complexity for a one-liner guard. Happy to add one if preferred.
Release Notes:
not signed inerror log on every keystroke when not signed in to Zed