Repository navigation
fix(s2s): start the pipeline once under the application - #60
WingedGuardian wants to merge 1 commit into
Conversation
Build and return the pipeline without scheduling another runner invocation. Keep startup, failure reporting and cleanup in Application.run; adapt the real-pipeline test host and pin the one-start regression.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the
 |
|
@codex review |
|
Codex Review: Didn't find any major issues. You're on a roll. Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
What changes
The application starts the S2S pipeline once. Previously
build_pipeline()scheduledrunner.run(task)in the background andApplication.run()awaited the same runner/task again, producing duplicate starts.The builder now constructs and returns the pipeline;
Application.run()owns execution, failure reporting and final cleanup. The obsolete background-task callback is removed. The real-pipeline transcript test is updated to start its returned runner explicitly.Closes #50.
Depends on #59; this is PR 2 of the three-PR reliability stack. Base:
fix/s2s-user-transcript-disconnect. Merge #59 first, then retarget this PR to main and refresh CI before merging.Validation
E2E: after merge and an owner-approved deploy, restart the bridge through its normal deployment path and verify one pipeline startup with no duplicate-worker error; confirm a voice conversation and clean shutdown still work.