Repository navigation
Release: Prerelease 10.6.0-alpha.2 - #35523
Merged
Merged
Conversation
The web-components examples in the loaders documentation were calling `TodoItem` as a function without ever importing or defining it — a copy-paste error from the React/Solid versions. Replaced with the correct `html` tagged template literal from `lit`, consistent with all other web-components snippets in the repository. Co-authored-by: Cursor <cursoragent@cursor.com>
The standalone (build) job per template was ~90s of which ~55s was fixed
setup (spin-up, checkout, workspace attach and unpack) around a ~25-45s
build - and it sat between (create) and the e2e/chromatic/vitest jobs on
the workflow critical path, adding a full job hop plus a workspace
round-trip to those chains.
The static build now runs at the end of the create job, before the sandbox
is packed, so storybook-static travels inside the existing sandbox tarball
and downstream jobs need no extra restore step. e2e, chromatic, vitest and
test-runner depend on create directly.
The Windows sandbox jobs and the daily test-runner job referenced flow
jobs by array index; they now use named references, since the build job no
longer exists. saveBench('build') records identical data - it runs inside
'yarn task build' regardless of which CI job hosts it, and no job in the
generated workflows consumed the bench task output.
…nary storybook init hardcodes skipInstall: true for addon postinstall scripts to keep dependency handling to itself, but both the addon-vitest and addon-a11y postinstalls also derived the package-runner mode from that flag - so every init paid for ephemeral dlx/npx environments fetching a storybook copy that was already installed as a devDependency one step earlier. A new useRemotePkg postinstall option decouples the runner choice from the subprocess flags: init passes the user's actual --skip-install value, so the locally installed binary runs via exec whenever it exists, while a genuinely skipped install still falls back to the ephemeral runner. The a11y postinstall is also skipped when addon-vitest was configured in the same init run: its single automigration (addon-a11y-addon-test) is already executed by vitest's postinstall when both addons are present, so the second invocation only spun up another package runner to conclude there was nothing to do. It still runs standalone (storybook add) or when vitest fails.
…h the sandbox The Angular dev and e2e jobs boot storybook through a docs:json prefix that unconditionally regenerates documentation.json, stalling the dev-server start behind a full Compodoc pass. On CI the sandbox tree is immutable once the create job packs it, so the documentation.json travelling in the sandbox archive is current by construction and the second run only reproduces it. Local sandboxes keep regenerating on every start since sources change there.
After the build-in-create, local-postinstall and Compodoc-skip changes the merged workflow tail is a tight cluster of dev-mode e2e jobs (~685s). Those jobs run a dev server and Playwright in parallel on a large (4 vCPU) executor with 3 workers. xlarge gives 8 vCPUs: ~2 for the dev server and 6 for the test workers, keeping the headroom ratio that made large/3 the sweet spot when the class was last benchmarked. Only the e2e-mode dev jobs change (17 in merged, 31 in daily); the smoke-only dev jobs stay on medium classic and the serve-based e2e jobs stay on medium+ with 3 workers. Dev-job minutes cost double (20 -> 40 credits), traded deliberately for wall clock on the critical tail.
The full-fleet xlarge/6-workers experiment cut the dev test phases 38% and the matched-pair wall by ~55s, but at +45% credits on the whole e2e-dev family (~+750k credits/30d fleet-wide) - more than the rest of the CI stack saves. The wall only moves when the slowest chain moves, so only the templates that define the tail keep the big class: angular-cli (webpack and vite), next-js latest (webpack and vite), react (vite and webpack 18) and vue3 vite. The other dev jobs return to large with 3 workers, where extra speed cannot move the wall. react-webpack/18-ts is included beyond the obvious tail set: in the matched morning baseline it was the second-latest finisher (334s duration), so leaving it on large would simply make it the new wall.
…build CI: Build sandbox static output inside the create job
…tall-automigrate CLI: Run addon postinstall automigrations with the local storybook binary
…oc-skip CI: Skip the redundant Compodoc run when documentation.json ships with the sandbox
Review feedback: defineSandboxJob_dev looked up the template in XLARGE_DEV_TEMPLATES itself; the call site now passes resourceClass explicitly, which keeps the job factory free of template knowledge. Also drop the PR reference from the template-set comment in favor of the durable selection criterion. Generated configs are byte-identical.
…kers CI: Run the tail-defining Playwright dev jobs on xlarge with more workers
Co-authored-by: Steve Dodier-Lazaro <Sidnioulz@users.noreply.github.com>
Co-authored-by: Steve Dodier-Lazaro <Sidnioulz@users.noreply.github.com>
Co-authored-by: Steve Dodier-Lazaro <Sidnioulz@users.noreply.github.com>
github-actions
Bot
added
the
release
The publish workflow's `Label patch PRs as picked` step uses the GraphQL `addLabelsToLabelable` mutation, which requires `pull-requests: write`. Commit 93fb631 swapped this permission to `issues: write`, so the `patch:done` labeling has failed with "Resource not accessible by integration" since 10.5.0 reached latest-release. The failure was silent (the script swallows the error), so already-released patch PRs kept being re-picked into subsequent patch changelogs. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…angelog-a4830f CI: Restore `pull-requests:write` permission for patch labeling
…ell-component TanStack: Keep Start's document shell out of stories
storybook-bot
force-pushed
the
version-non-patch-from-10.6.0-alpha.1
branch
from
July 17, 2026 13:03
1ae7a70 to
dbc107c
Compare
`label-patches.ts` has caught and discarded labeling errors since it was introduced in 2023 (#22959). That was fine for its original manual CLI use (the ora spinner shows a human the failure), but the same run() is invoked unattended by the publish workflow, where a swallowed error is invisible: the step exits 0 and the workflow stays green. That is how the recent pull-requests:write regression went unnoticed for two patch releases. - label-patches.ts now re-throws after logging, so the process exits non-zero. - The publish workflow marks the step continue-on-error (labeling is bookkeeping for the next release's cherry-pick and must not abort an already-published release) and alerts to Discord on failure. - Adds a regression test asserting the error propagates. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Keep the change to the behavior fix only. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…nt-failure Fix(ci): surface patch-labeling failures instead of swallowing them
storybook-bot
force-pushed
the
version-non-patch-from-10.6.0-alpha.1
branch
from
July 17, 2026 13:45
dbc107c to
1b08db6
Compare
…ocs-navigation-type Search: Show docs stories as docs search result item instead of a story-group search result item
storybook-bot
force-pushed
the
version-non-patch-from-10.6.0-alpha.1
branch
from
July 17, 2026 14:38
1b08db6 to
61ac103
Compare
…er-snippet Docs: Fix broken web-components snippets in loaders documentation
storybook-bot
force-pushed
the
version-non-patch-from-10.6.0-alpha.1
branch
from
July 17, 2026 15:48
61ac103 to
8f6c90b
Compare
ndelangen
added
ci:daily
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
8 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.
This is an automated pull request that bumps the version from `10.6.0-alpha.1` to `10.6.0-alpha.2`.
Once this pull request is merged, it will trigger a new release of version `10.6.0-alpha.2`.
If you're not a core maintainer with permissions to release you can ignore this pull request.
To do
Before merging the PR, there are a few QA steps to go through:
And for each change below:
This is a list of all the PRs merged and commits pushed directly to `next`, that will be part of this release:
If you've made any changes doing the above QA (change PR titles, revert PRs), manually trigger a re-generation of this PR with this workflow and wait for it to finish. It will wipe your progress in this to do, which is expected.
Feel free to manually commit any changes necessary to this branch after you've done the last re-generation, following the Make Manual Changes section in the docs, especially if you're making changes to the changelog.
When everything above is done:
Generated changelog
10.6.0-alpha.2