Repository navigation
✨ Repeat browser workflows with reproducible seeds - #33
Merged
Merged
Conversation
Exercise complete selections within one managed session, retiring final documents before project draining and advancing only after persistence. Keep failures visible across rounds and save bounded replay evidence, aggregate reports, and source-generated guides.
Preserve explicit execution options in replay commands, retain accepted outcomes through artifact write failures, and publish terminal batch metadata last. Keep automatic batch replacement under one owner and give repeated browser workflows their own CI budget without dropping engine coverage.
Document the source and build policies needed by the pinned Interactor fixes so projects can make that installation choice explicitly.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Acceptance tests can pass once while depending on test order or state left by earlier workflows.
dyno --suite Checkout --repeat 100 --randomexercises complete selections in one reused managed session, preserves every observed failure, and prints commands to replay a failing round or rerun its prefix.Each round remains an ordinary Dyno run. Final documents park before project draining, advancement waits for cleanup and persistence, and each round has its own deadline. Backend reset stays project-owned, and Interactor owns convergence. Seeds reproduce priority; schedules reproduce worker queues, not asynchronous interleaving. UI, manual and watch combinations are rejected explicitly.
Batch HTML, JSON, JUnit and schedules retain failures independently of live history. Successful intermediate screenshots are discarded; failed, interrupted and final rounds keep portable reports within a 256 MiB evidence budget. Replay commands retain config, browser and report overrides. Storage errors preserve accepted outcomes when the remaining output paths are writable, and terminal metadata is published after the final artifacts. Reporter reruns cannot bypass automatic batch ownership.
The complete local framework suite passed with 262 passes and two skips. Real CLI coverage includes resource reuse, seeded replay, storage failures, cleanup timeouts, all-skipped selections, signals and offline reports. Focused Chromium and WebKit checks passed, as did the docs build and seven docs tests. Repeat execution has a separate CI suite so all browser engines stay within the existing job limit. Local Firefox could not open its installed profile; the CI matrix verifies Firefox independently.
Consumer archive checks passed Vizzly’s 355 workflows, BackChannel’s 31 desktop/mobile workflows, and the docs app’s 40 workflows. Installation guidance also calls out pnpm 11’s Git dependency policy, which must be an explicit project choice while the required Interactor fixes remain unreleased.