Skip to main content

Release Notes

These notes cover changes that affect package users from 18 October 2025 to 18 October 2026. The newest release appears first. Releases with only internal maintenance changes are omitted. Each published entry links to its GitHub release. Older history is available in all GitHub releases.

1.1.118 - Released​

Version 1.1.118 is published and available.

Added​

  • Reporter and CLI configurations support artifactLinks for hosted screenshots, traces, other attachments, and HTML reports beside failures.
  • Bot reports include these links in failure threads when enabled. Webhook reports show links inline.
  • Custom layouts receive optional failures[].artifacts entries with names, URLs, and content types when available.
  • The failure artifacts template provides a preview, downloadable TypeScript and Block Kit JSON, and setup examples.
  • Attachment paths outside the configured upload directory are omitted. Oversized links are omitted with a notice in default layouts.

Fixed​

  • CLI prechecks reject a missing customLayoutAsync.source file before loading the report.
  • Failure details keep same-named tests separate across files, projects, and repetitions, so each failure retains its own artifacts.
  • The checked-in custom_block/simple.ts example no longer reports success for unsuccessful runs or global errors. If you use a copied simple layout, include runStatus and runErrors in its summary checks.
  • CLI prechecks validate that configured customLayoutAsync.source files exist before checking credentials or sending reports.

Upgrade instructions​

Upload the artifacts to your host with their directory structure intact. Set artifactLinks.baseUrl to the uploaded directory URL. Set artifactLinks.pathPrefix to the corresponding local directory. To add the HTML report, set artifactLinks.reportUrl to its hosted URL.

The package does not upload files or change access permissions. Existing configurations require no changes. Links use the final failing attempt's attachments and follow the failure display limit. Reporter and CLI option names are identical. CLI paths resolve from its working directory.

Reporter options · CLI options

1.1.117 - 11 October 2026​

Added​

  • Slack reports show flaky test names, projects, and actual retry counts. Each test appears once per project and repetition.
  • The reporter and CLI support sendResults: 'on-flaky'. This sends a report only when the run contains flaky tests.
  • Flaky details have a separate display limit. The reporter uses maxNumberOfFlakyTestsToShow; the CLI uses maxNumberOfFlakyTests. Both default to 10.
  • Custom layouts receive the complete optional flakyTests array, including test names, files, projects, and retry counts.

Upgrade instructions​

For alerts about flaky tests, set sendResults to on-flaky. This setting skips runs with failures but no flaky tests. The existing on-failure setting still skips runs with only flaky tests.

To hide flaky details, set the applicable display limit to 0. The flaky count and notification rules remain active.

For bot reports with threaded details, set showInThread to true. Webhook reports show details in the main message.

Reporter options · CLI options · GitHub release

1.1.116 - 10 October 2026​

Fixed​

  • Failure notifications include global errors, timeouts, and graceful interruptions, even when no individual test failed.
  • The default layout shows an unsuccessful run status. Global errors appear separately and do not increase the failed-test count.
  • Bot reports route unsuccessful runs to onFailureChannels when configured.

Added​

  • Custom layouts receive the optional runStatus and runErrors fields.
  • The CLI accepts --run-status with passed, failed, timedout, or interrupted. Playwright JSON does not contain the overall run status.

Upgrade instructions​

If you use a custom layout, check runStatus and runErrors when you determine success. Do not use failed === 0 as the only success condition. The custom layout example includes these checks.

If your runner detects a global timeout, add --run-status timedout to the CLI command. Without this option, the CLI infers the status from available errors and test results.

Notifications require Playwright to complete the reporter callback. Forced process termination can prevent delivery.

Run errors · CLI run status · GitHub release

1.1.115 - 10 October 2026​

Added​

The documentation website provides setup, configuration, and troubleshooting guidance for package users. It includes webhook setup, bot setup, CLI usage, and custom layouts. No configuration change is required to use the documentation.

GitHub release

1.1.114 - 4 October 2026​

Fixed​

Bot delivery no longer produces unreadable Slack API responses caused by proxy dependency loading in version 1.1.113. The problem could affect connections without a proxy. The proxy dependency now loads only when a proxy is configured.

Upgrade instructions​

If version 1.1.113 produces unreadable Slack API responses, install version 1.1.114 or later. Keep your existing bot and proxy settings.

Installation · Fix details · GitHub release

1.1.113 - 22 August 2026​

Added​

Custom layouts can read Playwright test tags from the optional tags field on each SummaryResults.tests entry. Tags are available through both the reporter and the CLI. This supports layouts that group or identify tests by tag.

Upgrade instructions​

To display tags, read each test's tags field in your custom layout. If the field is absent, use an empty array.

Custom layouts · GitHub release

1.1.110 - 14 April 2026​

Fixed​

The Slack client dependency uses patched Axios versions to address reported security issues in its HTTP dependency. This change updates the dependencies installed with the package. It does not require changes to your Slack configuration.

Upgrade instructions​

If you use an earlier version, install version 1.1.110 or later. Update your project's lockfile with your package manager.

Installation · Dependency fix · GitHub release

1.1.106 - 14 February 2026​

Changed​

The package no longer installs the unused ts-node runtime dependency or the diff dependency through that path. This reduces the dependencies needed by package users. Other packages in your project can still depend on those libraries. Your reporter and CLI configuration can stay the same.

Dependency change · GitHub release