Complete Azure DevOps live-published test results

[Amaury Levé (Evangelink)]

Azure DevOps keeps live-published test results in Pending when the payload contains a terminal outcome but omits the result-level state. This change publishes state: "Completed" for every terminal result so completed runs report accurate passed, failed, and incomplete counts.

The regression coverage verifies:

• MTP-to-Azure DevOps mapping for passed, failed, skipped, timed out, and cancelled results.
• The actual initial POST payload through the browser-WASM fake Azure DevOps REST endpoint.
• The initial POST and rerun PATCH payload contracts, including the separate PATCH serializer configuration.

The adjacent-path audit confirmed retry attempts use the same parent result model. Azure DevOps TestSubResult has no result-state field, so no equivalent sub-result change is required.

Fixes: #11122

[Copilot]

Copilot review overview

🟢 Approval recommended

The fix and regression coverage are complete, with no unresolved issues.

Review tier: Balanced
Findings: None

What changed in this PR

Fixes Azure DevOps live-published terminal test results remaining Pending by explicitly setting their state to Completed.

Changes:

• Adds state: "Completed" to terminal results.
• Adds unit and browser-WASM regression coverage.
• Updates the internal API baseline.

File	Description
test/​UnitTests/​Microsoft.Testing.Extensions.UnitTests/​AzureDevOpsLivePublishingTests.cs	Verifies terminal-state mapping.
test/​IntegrationTests/​Microsoft.Testing.Platform.Acceptance.IntegrationTests/​BrowserWasmExecutionTests.cs	Verifies serialized REST payload states.
src/​Platform/​Microsoft.Testing.Extensions.AzureDevOpsReport/​InternalAPI/​InternalAPI.Unshipped.txt	Tracks the new internal getter.
src/​Platform/​Microsoft.Testing.Extensions.AzureDevOpsReport/​AzureDevOpsLivePublishingModels.cs	Adds the completed result state.

---

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

[Copilot]

Copilot review overview

🟢 Approval recommended

The completed state is consistently implemented and covered by unit and integration tests.

Review tier: Balanced
Findings: None

[github-actions]

Pipeline Test Triage — Final Resolution

Build: [1588618]((dev.azure.com/redacted) (microsoft.testfx, full/completed analysis)
PR: #11142 — head 3cf82ae0 (matches expected head/merge SHAs)

Result: No actionable test failures found.

• All 90,841 collected test records across CTRF/TRX/JUnit reports for this build are passed or skipped (400 skipped, all with expected reasons such as OS-restriction or environment gating). failureOrRetryCount = 0, diagnosticCount = 0, slowRegressionCount = 0 (37 candidates flagged but none met the 10-sample/60s-floor/3x-p95 slowness threshold). Evidence is complete (evidenceIncomplete: false).
• The pipeline's overall failed result stems from the MacOS Release job being cancelled due to a worker timeout (Job cancelled due to worker timeout, Publish Test Results folders step also cancelled), not from any test assertion failure. The Test task itself completed successfully before the cancellation, and the pipeline timeline explicitly notes: "macOS is non-blocking for PRs; this failure does not block the merge."
• This is an infrastructure/agent-timeout issue rather than a test-product defect, and is out of scope for this test-triage agent (ordinary build/agent failures are handled separately). No recurring or historical pattern for a specific test was identified since there is no failing test to correlate.

Conclusion: No issue created — no durable test defect, flaky test, crash, or slowness regression threshold was met. The macOS job cancellation is a non-blocking infrastructure timeout; if it recurs across multiple builds, it would warrant separate build-infrastructure investigation rather than a test-triage issue.

🤖 Automated content by GitHub Copilot. Generated by the Pipeline Test Triage workflow. · auto · 36.3 AIC · ⌖ 1.78 AIC · ⊞ 9.3K · [◷]( · ◷)