Repository navigation
fix(inference): discover paired hosts and refresh routed models - #151
Merged
ScriptedAlchemy merged 2 commits intoNov 18, 2026
Merged
Conversation
|
ScriptedAlchemy
changed the base branch from
codex/blur-inference-account-names
to
main
November 18, 2026 03:39
ScriptedAlchemy
enabled auto-merge
November 18, 2026 03:39
This was referenced Nov 18, 2026
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
T3 could keep an old inference host catalog and classify newly advertised Codex models as custom. That left GPT-6.2 Sol and Astra out of the model picker. Enabling routing could also send users into a separate settings page to configure a host they had already paired.
The provider switch now discovers eligible local and paired hosts, reuses existing access, and offers account setup inline when no host is ready. Sharing and revocation remain under host control. Web, desktop, and mobile use the same host selection logic. Provider refresh updates the selected route and its models, including project-enabled providers and fallback hosts. Codex and Claude expose advertised chat models to the picker.
The dead-code review removed unused provider fields, a redundant route check, and an unnecessary host probe. It also turned up stale setup state when turning routing off or switching providers, now covered by browser regressions.
The account-name blur is in #150. This PR is stacked on that branch.
Verified with changed-file checks, focused server and browser tests, and web and mobile typechecks. An isolated browser scenario reproduced the missing setup action on main and passed with the fix. A second scenario verified that refreshing Codex exposes both GPT-6.2 models after the host advertises them.
Built with GPT-6 in Codex Desktop.