Add eval coverage for dotnet-test/grade-tests

[Evangelink]

Extends tests/dotnet-test/grade-tests/eval.yaml with two new scenarios (plus supporting fixtures) so the previously-uncovered teaching points in plugins/dotnet-test/skills/grade-tests/SKILL.md are now exercised.

Now-covered points

• [Validation] Every grade is justified by at least one observable signal (L307)
• [Pitfall] Inflating deductions to justify the grade (L331)
• [Pitfall] Flagging Go/Rust table-driven loops as conditional logic (L335)
• [Pitfall] Penalizing tests when production code is unavailable (L337)
• [Pitfall] Spilling a 500-row table into a PR comment (L339)

What was added

• Scenario 4 — Go table-driven tests (fixtures/go-table-driven/): grades idiomatic for ... range / t.Run subtests; rubric asserts the loop is NOT misread as conditional logic, every grade rests on an observable signal, and deductions are not inflated.
• Scenario 5 — production code unavailable (fixtures/production-unavailable/): grades tests whose code under test (Payments.Core) is absent; rubric asserts behavioral concerns are marked Unverified (not deducted) and the report stays compact rather than spilling a giant 500-row table into the PR comment.

Prompts are natural (no skill references); rubric items are outcome-focused and independently evaluable.

Verification

• Measure-SkillCoverage.ps1 -PluginName dotnet-test -SkillName grade-tests → 100% (28/28), uncovered empty, no regressions.
• SkillValidator check --plugin ./plugins/dotnet-test → ✅ all checks passed (YAML parses; remaining warnings are pre-existing token-size notes).

No SKILL.md or other skills were modified.

[github-actions]

Skill Coverage Report

	Plugin	Skill	Covered	Coverage
✅	dotnet-test	grade-tests	28/28	100%

[Copilot]

Pull request overview

This PR expands the dotnet-test/grade-tests evaluation suite by adding two new scenarios (Go table-driven tests; production code unavailable) plus new fixtures to exercise previously uncovered guidance in the grade-tests skill.

Changes:

• Added Scenario 4 to grade idiomatic Go table-driven tests and ensure loops/assertion patterns aren’t misclassified as branching.
• Added Scenario 5 to grade tests when the production code is missing, expecting “Unverified” (not deductions) and compact PR-comment output.
• Added supporting Go and C# fixture workspaces for the new scenarios.

Show a summary per file

File	Description
tests/dotnet-test/grade-tests/eval.yaml	Adds two new eval scenarios and their assertions/rubrics to cover additional grade-tests teaching points.
tests/dotnet-test/grade-tests/fixtures/go-table-driven/go.mod	Introduces a minimal Go module for the table-driven test fixture.
tests/dotnet-test/grade-tests/fixtures/go-table-driven/calculator.go	Adds simple Go production code to be graded against by the Go scenario.
tests/dotnet-test/grade-tests/fixtures/go-table-driven/calculator_test.go	Adds Go tests of varying quality (A/C/F) including a table-driven subtest loop.
tests/dotnet-test/grade-tests/fixtures/production-unavailable/Payments.Tests/Payments.Tests.csproj	Adds a .NET test project fixture explicitly lacking the production project reference.
tests/dotnet-test/grade-tests/fixtures/production-unavailable/Payments.Tests/PaymentGatewayTests.cs	Adds MSTest methods of varying quality for the “production unavailable” grading scenario.

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

• Files reviewed: 6/6 changed files
• Comments generated: 2

[Evangelink]

/evaluate

[github-actions]

Skill Validation Results

Skill	Scenario	Quality	Skills Loaded	Overfit	Verdict
grade-tests	Grade a curated list of MSTest tests with mixed quality	4.3/5 → 5.0/5 🟢	✅ grade-tests; tools: skill, glob / ✅ grade-tests; tools: skill, glob, bash, task, read_agent	✅ 0.20	❌ [1]
grade-tests	Grade pytest test methods using the same rubric	5.0/5 → 4.0/5 🔴	✅ grade-tests; tools: skill, glob, bash / ✅ grade-tests; tools: skill, glob	✅ 0.20	❌ [2]
grade-tests	Decline to grade the entire workspace when no test list is provided	1.0/5 → 5.0/5 🟢	✅ grade-tests; tools: skill / ✅ grade-tests; test-analysis-extensions; tools: skill, bash, glob	✅ 0.20	❌ [3]
grade-tests	Grade Go table-driven tests without misreading the loop as branching	4.0/5 → 4.0/5	✅ grade-tests; tools: skill / ✅ grade-tests; tools: skill, glob, bash	✅ 0.20	❌ [4]
grade-tests	Grade tests when the production code under test is unavailable	4.0/5 → 4.0/5	✅ grade-tests; tools: skill, glob, bash / ✅ grade-tests; tools: skill, bash, glob	✅ 0.20	❌ [5]

[1] ⚠️ High run-to-run variance (CV=318%) — consider re-running with --runs 5. (Plugin) Quality unchanged but weighted score is -6.2% due to: tokens (40505 → 156555), tool calls (3 → 8), time (51.0s → 156.7s)
[2] (Plugin) Quality unchanged but weighted score is -8.6% due to: tokens (26417 → 122036), tool calls (2 → 6), time (28.0s → 112.7s)
[3] (Plugin) Quality unchanged but weighted score is -12.7% due to: errors (0 → 1), tokens (52678 → 105055), time (71.9s → 118.6s), tool calls (7 → 10)
[4] ⚠️ High run-to-run variance (CV=133%) — consider re-running with --runs 5
[5] (Plugin) Quality unchanged but weighted score is -9.2% due to: tokens (26558 → 131127), tool calls (2 → 7), time (47.0s → 110.8s)

⏰ timeout — run(s) hit the (120s) scenario timeout limit; scoring may be impacted by aborting model execution before it could produce its full output (increase via timeout in eval.yaml)

Model: claude-opus-4.6 | Judge: claude-opus-4.6

🔍 Full Results - additional metrics and failure investigation steps

To investigate failures, paste this to your AI coding agent:

For PR 825 in dotnet/skills, download eval artifacts with gh run download 28157445404 --repo dotnet/skills --pattern "skill-validator-results-*" --dir ./eval-results, then fetch https://raw.githubusercontent.com/dotnet/skills/6b6633b6c01c3ca134e0a53af5e52365ca699461/eng/skill-validator/src/docs/InvestigatingResults.md and follow it to analyze the results.json files. Diagnose each failure, suggest fixes to the eval.yaml and skill content, and tell me what to fix first.

[github-actions]

👋 @Evangelink — this PR has 1 unresolved review thread(s). When you're ready, please address the feedback and push an update; the triage bot will pick up the next state automatically. (Add the no-stale label to silence further pings.)

[github-actions]

✅ Evaluation passed for 6b6633b. cc @dotnet/dotnet-testing — please review.