CoR: Add telemetry for azure-deploy artifacts

[Matthew Fisher (MicroFish91)]

No description provided.

[Copilot]

🟡 Changes recommended

Outstanding issues affect telemetry timing, failure-state reporting, and Deployment Plan data compatibility.

Get a fresh assessment by requesting another Copilot review.

Pull request overview

Adds privacy-filtered telemetry for azure-deploy artifacts during deployment-result handoff.

Changes:

• Adds artifact parsing and allowlisted telemetry metrics.
• Adds fixtures and focused parsing, failure, privacy, and integration tests.
• Wires telemetry into the deployment-results handoff.

File summaries

File	Summary
test/testProjects/copilotOnRails/attendance/scaffold-manifest.json	Scaffold telemetry fixture
test/testProjects/copilotOnRails/attendance/prereq-output.json	Prerequisite telemetry fixture
test/testProjects/copilotOnRails/attendance/prepare-plan.json	Needs parser-compatible service and cost fields restored
test/testProjects/copilotOnRails/attendance/deploy-result.json	Deployment telemetry fixture
test/copilotOnRails/recordDeployArtifactsTelemetry.test.ts	End-to-end recording tests
test/copilotOnRails/deployScaffoldTelemetryUtils.test.ts	Scaffold extraction tests
test/copilotOnRails/deployResultTelemetryUtils.test.ts	Deployment extraction tests
test/copilotOnRails/deployPrereqTelemetryUtils.test.ts	Prerequisite extraction tests
test/copilotOnRails/deployPrepareTelemetryUtils.test.ts	Preparation extraction tests
test/copilotOnRails/deploymentPlanTelemetryUtils.test.ts	Deployment-plan telemetry coverage
src/webviews/copilotOnRails/extension/utils/deployArtifactTelemetryUtils.ts	Needs failure-state metric emission and documentation updates
src/webviews/copilotOnRails/extension/openDeployResultView.ts	Returns the selected result URI
src/chat/tools/copilotOnRails/openDeployResultViewTool.ts	Needs awaited inventory capture and diagnostics documentation updates

Review details

Suppressed comments (4)

src/chat/tools/copilotOnRails/openDeployResultViewTool.ts:30

• The new call records four artifact-derived properties into the workspace-cached diagnostics via setCorProp, but docs/copilot-create-project.md still describes only the generic properties bag and does not document these new fields or their allowlisted privacy contract. Update the support/triage diagnostics and privacy section in the same change so Inspect Diagnostics and Report Issue remain accurate.

                    await recordDeployArtifactsTelemetry(corContext, resultFile);

src/webviews/copilotOnRails/extension/utils/deployArtifactTelemetryUtils.ts:112

• Returning before setCorProp when deployResult.status is not terminal drops every collected metric, including deployResult.parsedOk: false or readFailed for a malformed/unreadable final artifact. Those are the diagnostics needed for exactly that failure mode; emit the allowlisted properties even when the result is missing, invalid, or unfinished instead of discarding them.

    if (properties['deployResult.status'] !== 'succeeded' && properties['deployResult.status'] !== 'failed') {
        return;

test/testProjects/copilotOnRails/attendance/prepare-plan.json:43

• This fixture no longer matches the fields consumed by parsePreparePlanJson: its service entries use id/kind/componentId, while the parser reads name/sku/component/resourceName and drops entries that have neither name nor resourceName. Opening this project's Deployment Plan therefore renders missingServices; removing the attendance deployment-plan test hides the regression. Keep the canonical fields or update the parser and add a view regression test.

      "id": "attendance-api",
      "kind": "azure-functions",
      "sku": "Flex Consumption (FC1)",
      "region": "eastus2",
      "runtime": "node",

test/testProjects/copilotOnRails/attendance/prepare-plan.json:190

• The Deployment Plan parser only reads costEstimate.monthlyUsd and costEstimate.breakdown[], but this fixture replaces them with monthlyTotalUsd and items. It will display a $0 estimate with no breakdown in the Deployment Plan, and the removed attendance test no longer catches it. Preserve the canonical fields or teach the parser this alternate layout.

    "monthlyTotalUsd": 32,
    "monthlyRangeUsd": {"low": 25, "high": 40},
    "items": [
      {"service": "Function App (Flex Consumption FC1)", "monthlyUsd": 2, "assumption": "Dev traffic: ~500K executions, mostly under free grant."},
      {"service": "Storage Account (Standard_LRS)", "monthlyUsd": 1, "assumption": "Functions runtime + small blob usage."},

• Files reviewed: 13/13 changed files
• Comments generated: 2
• Review effort level: Lite

---

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