Emit queue: max on top-level workflow concurrency groups

[Copilot]

The compiler already emits queue: max by default on job-level concurrency groups (agent, output, conclusion), but the top-level workflow concurrency group was left at the Actions default (queue: single). This meant back-to-back triggers — most notably rapid pushes — could displace a queued run before it ever executed, silently dropping work instead of queuing it.

Compiler change

• GenerateConcurrencyConfig (pkg/workflow/concurrency.go) now appends queue: max to the generated top-level group whenever cancel-in-progress is not enabled for that trigger (the two are mutually exclusive per GitHub Actions).
• Respects the existing features.group-concurrency-queue: false opt-out via isGroupConcurrencyQueueEnabled, consistent with job-level groups.

concurrency:
  group: "gh-aw-${{ github.workflow }}-${{ github.ref || github.run_id }}"
  queue: max

Tests & fixtures

• Updated pkg/workflow/concurrency_test.go expectations for all non-cancelling trigger cases.
• Regenerated wasm golden fixtures.

Generated workflows

• Recompiled all .github/workflows/*.md sources so .lock.yml files reflect the new default (each gains a single queue: max line under its top-level concurrency block; PR-triggered workflows are unaffected since they use cancel-in-progress: true).

Docs

• Updated docs/src/content/docs/reference/concurrency.md to describe the new default and the opt-out.

• Fixes Top-level workflow concurrency group omits queue: max, so push-triggered workflows silently drop queued runs #52547

[Copilot]

Pull request overview

Updates compiler-generated workflow concurrency so non-cancelling runs queue sequentially instead of displacing pending runs.

Changes:

• Emits top-level queue: max unless cancellation or the feature opt-out is active.
• Updates concurrency tests, documentation, and golden fixtures.
• Recompiles generated workflow lock files.

Show a summary per file

File	Description
pkg/workflow/concurrency.go	Generates top-level queue configuration.
pkg/workflow/concurrency_test.go	Updates concurrency expectations.
docs/src/content/docs/reference/concurrency.md	Documents defaults and opt-out.
pkg/workflow/testdata/TestWasmGolden_CompileFixtures/with-imports.golden	Updates golden output.
pkg/workflow/testdata/TestWasmGolden_CompileFixtures/playwright-cli-mode.golden	Updates golden output.
pkg/workflow/testdata/TestWasmGolden_CompileFixtures/basic-copilot.golden	Updates golden output.
pkg/workflow/testdata/TestWasmGolden_AllEngines/pi.golden	Updates engine golden output.
pkg/workflow/testdata/TestWasmGolden_AllEngines/gemini.golden	Updates engine golden output.
pkg/workflow/testdata/TestWasmGolden_AllEngines/copilot.golden	Updates engine golden output.
pkg/workflow/testdata/TestWasmGolden_AllEngines/codex.golden	Updates engine golden output.
pkg/workflow/testdata/TestWasmGolden_AllEngines/claude.golden	Updates engine golden output.
.github/workflows/*.lock.yml (273 files)	Regenerates workflows with top-level queue: max.

Review details

💡 Add a code-review agent skill for context-aware, tailored reviews. Learn more in the docs.

• Files reviewed: 283/283 changed files
• Comments generated: 1
• Review effort level: Balanced

[pelikhan]

@copilot make sure there is a way to turn off Q-Max for that frontmatter.

[Copilot]

@copilot make sure there is a way to turn off Q-Max for that frontmatter.

The opt-out already applies to the top-level group: GenerateConcurrencyConfig calls the same isGroupConcurrencyQueueEnabled check used for job-level groups, so features.group-concurrency-queue: false disables queue: max everywhere, including the top-level workflow group. Added a test case (TestGenerateConcurrencyConfig) verifying this in 33a1f8d (see commit).