Generated banner for Test Analysis Extensions
Generated logo for Test Analysis Extensions

Test Analysis Extensions

.NET (Microsoft)OfficialPublished by .NET (Microsoft)

  • Skill
  • Meta (authoring, orchestration, planning, memory)
  • Agent tooling
  • MIT
  • No issues found

From the author

Provides file paths to language-specific reference files for the test ANALYSIS skills (assertion-quality, test-anti-patterns, test-gap-analysis, test-smell-detection, test-tagging). Read this catalog to discover available extension files (e.g., dotnet.md for .NET/MSTest/xUnit/NUnit/TUnit, python.md for pytest/unittest, typescript.md for Jest/Vitest/Mocha, java.md for JUnit/TestNG, etc.). Do not invoke this reference-only skill — its bundled files are read by the quality agent and analysis skills needing framework-specific lookup tables (test markers, assertion APIs, skip annotations, sleep patterns, mystery guest indicators, integration markers, setup/teardown, tag-support capability).
— the description by .NET (Microsoft)

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

Test Analysis Extensions is a non-invocable data repository designed for internal use by other test-related agent skills. It provides structured reference files for various programming languages and their respective testing frameworks. These files contain essential metadata for identifying test markers, assertion APIs, lifecycle hooks, and anti-pattern indicators, enabling analysis agents to perform language-neutral evaluations across diverse codebases.

What it does

  • Provides standardized lookup tables for test framework syntax and conventions.
  • Maps language-specific features like setup/teardown hooks and skip annotations.
  • Defines capability support levels for automated tagging and reporting.
  • Enables consistent detection of mystery guest indicators and integration markers.
  • Supports cross-language analysis by normalizing test-related concepts.

When to use it

  • When building or maintaining test analysis agents that require framework-specific knowledge.
  • When an agent needs to identify test lifecycle methods or assertion patterns in a specific language.
  • When determining if a test suite supports automated tagging or requires report-only auditing.

Advantages

  • Decouples analysis logic from framework-specific implementation details.
  • Ensures consistent detection patterns across multiple programming languages.
  • Provides clear capability definitions for automated test manipulation.
  • Facilitates easier extension of analysis agents to new languages or frameworks.

Usage examples

“N/A - This is a reference-only skill for internal agent use.”
The agent automatically reads the relevant language file (e.g., python.md) to identify pytest fixtures and markers during analysis.

Install

  1. Download the skill as a ZIP:

    https://autogineer.com/agent-skills/dotnet/test-analysis-extensions/download
  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip test-analysis-extensions.zip -d .agents/skills

Documentation by .NET (Microsoft)

Test Analysis Extensions

This reference-only catalog lists the bundled files used by test analysis. Consumers read the matching file directly; disable-model-invocation: true means this catalog must not be invoked through the model's skill tool. Resolve these resource paths relative to this catalog, not the project workspace.

Available Extension Files

FileLanguages / FrameworksContents
extensions/dotnet.md.NET (C#/F#/VB) — MSTest, xUnit, NUnit, TUnitTest markers, assertion APIs, sleep/delay patterns, skip annotations, mystery guest, integration markers, setup/teardown, tag support
extensions/python.mdPython — pytest, unittestSame categories, with pytest fixtures/markers and unittest TestCase
extensions/typescript.mdTypeScript / JavaScript — Jest, Vitest, Mocha, Jasmine, node:testSame categories, with async/await pitfalls
extensions/java.mdJava — JUnit 4, JUnit 5 (Jupiter), TestNGSame categories, with @Tag / @Category / groups
extensions/go.mdGo — testing package, testifySame categories, with table-driven idiom and build tags
extensions/ruby.mdRuby — RSpec, MinitestSame categories, with RSpec metadata and Minitest tags
extensions/rust.mdRust — built-in #[test], cargo testSame categories, with #[ignore], #[should_panic], feature flags
extensions/swift.mdSwift — XCTest, Swift TestingSame categories, with @Test, @Tag, @Suite
extensions/kotlin.mdKotlin — JUnit 5, Kotest, MockKSame categories, with @Tag and Kotest tags
extensions/powershell.mdPowerShell — Pester v5Same categories, with -Tag and Skip
extensions/cpp.mdC++ — GoogleTest, Catch2, doctestSame categories, with [tags] and * filters

Usage

  1. Detect the target codebase's primary language and test framework.

  2. Read the matching extension file before performing analysis.

  3. If multiple test frameworks are present (e.g., a project mixing Jest and Mocha), read all relevant extensions.

  4. Each extension file documents the same categories so analysis skills can be language-neutral.

Capability tags

Each extension file declares per-capability support so skills can gate behaviour safely:

  • Test discovery — how to locate test files and methods.

  • Assertion detection — framework-specific and language-level assertion forms.

  • Sleep/delay patterns — synchronous and asynchronous waits.

  • Skip / ignore — how to recognize skipped/ignored tests.

  • Setup / teardown — fixture and lifecycle hooks.

  • Mystery guest indicators — common file/db/network/env coupling patterns.

  • Integration markers — conventions that mark a test as integration/E2E.

  • Tag support (for test-tagging skill) — one of:

    • auto-edit — language has a canonical attribute/marker the skill can safely write.

    • report-only — no canonical syntax; produce audit reports without edits.

    • convention-based — tags exist via name/comment conventions only.

Notes for skill authors

  • Treat extension files as data, not as guidance to follow verbatim. They tell skills how to detect things in each language, not what to think about findings.

  • When language detection is uncertain, prefer reading multiple extension files over guessing.

  • If the user explicitly names a framework that does not have an extension file yet, fall back to the closest one (e.g., Pest → python.md/pytest semantics) and note the gap in the report.

Files

extensions/cpp.md

C++ Test Frameworks Reference (GoogleTest, Catch2, doctest, Boost.Test)

Reference data for analyzing C++ test code. Used by the polyglot test analysis skills.

Capability tags

CapabilitySupport
Test discoveryStrong — TEST* macros, TEST_CASE, DOCTEST_TEST_CASE
Assertion detectionStrong — ASSERT_*, EXPECT_*, REQUIRE, CHECK
Sleep/delay detectionStrong — std::this_thread::sleep_for, sleep(), Sleep()
Skip/ignore detectionModerate — GTEST_SKIP(), DISABLED_ prefix, [!hide] tags
Setup/teardown detectionStrong — SetUp/TearDown, fixtures, sections
Tag supportauto-edit — Catch2 uses [tag] syntax inside TEST_CASE; doctest uses * doctest::test_suite("tag") decorator chains; GoogleTest uses test-name prefix conventions (treat as convention-based)

Test File Identification

FrameworkFile conventionTest method markers
GoogleTest*_test.cc/cpp, *Tests.cppTEST(SuiteName, TestName), TEST_F(FixtureClass, TestName), TEST_P(...) parametrized, TYPED_TEST(...)
Catch2*Tests.cpp, test*.cppTEST_CASE("name", "[tags]"), SCENARIO, SECTION
doctest*Tests.cppTEST_CASE("name" * doctest::test_suite("suite"))
Boost.Test*_test.cppBOOST_AUTO_TEST_CASE(name), BOOST_FIXTURE_TEST_CASE(name, Fixture)

Assertion APIs

CategoryGoogleTestCatch2doctestBoost.Test
Equality (continue)EXPECT_EQ(actual, expected)CHECK(actual == expected)CHECK(actual == expected)BOOST_CHECK_EQUAL(actual, expected)
Equality (abort)ASSERT_EQ(actual, expected)REQUIRE(actual == expected)REQUIRE(actual == expected)BOOST_REQUIRE_EQUAL(actual, expected)
BooleanEXPECT_TRUE(x) / EXPECT_FALSE(x)CHECK(x) / CHECK_FALSE(x)CHECK(x)BOOST_CHECK(x)
Null/PointerEXPECT_EQ(ptr, nullptr)CHECK(ptr == nullptr)CHECK(ptr == nullptr)BOOST_CHECK(ptr == nullptr)
ThrowsEXPECT_THROW(stmt, ExType) / EXPECT_THROW(stmt, std::exception)CHECK_THROWS_AS(expr, ExType) / CHECK_THROWS_WITH(expr, "...") / CHECK_THROWS_MATCHES(...)CHECK_THROWS_AS(expr, ExType)BOOST_CHECK_THROW(expr, ExType)
No throwEXPECT_NO_THROW(stmt)CHECK_NOTHROW(expr)CHECK_NOTHROW(expr)BOOST_CHECK_NO_THROW(expr)
ApproximateEXPECT_NEAR(a, b, abs_err) / EXPECT_DOUBLE_EQ(a, b)CHECK(actual == Approx(expected))CHECK(actual == doctest::Approx(expected))BOOST_CHECK_CLOSE(a, b, tol_pct)
StringEXPECT_STREQ(c_str_a, c_str_b) / EXPECT_THAT(s, HasSubstr("x"))CHECK(s.find("x") != std::string::npos)similarBOOST_CHECK_EQUAL(s, expected)
Death testsEXPECT_DEATH(stmt, "regex") / EXPECT_EXIT(...)n/an/an/a
Custom matchersEXPECT_THAT(value, gmock_matchers::Eq(x))REQUIRE_THAT(value, Catch::Matchers::Equals(x))similarn/a

EXPECT vs ASSERT/REQUIRE vs CHECK:

  • GoogleTest: EXPECT_* continues on failure; ASSERT_* aborts the test.

  • Catch2 / doctest: CHECK* continues; REQUIRE* aborts.

  • Boost.Test: BOOST_CHECK* continues; BOOST_REQUIRE* aborts.

Sleep/Delay Patterns

PatternExample
C++11 thread sleepstd::this_thread::sleep_for(std::chrono::seconds(1))
POSIX sleepsleep(1); / usleep(500000);
Windows sleepSleep(1000);
Loop waitwhile (!ready) std::this_thread::sleep_for(...)
Async wait (acceptable)future.wait_for(std::chrono::seconds(5))

Skip/Ignore Annotations

FrameworkMechanism
GoogleTestGTEST_SKIP() << "reason"; inside test body; test name prefix DISABLED_ (e.g., TEST(F, DISABLED_X))
Catch2[!hide] or [.] tag in TEST_CASE("name", "[.]"); SUCCEED("skipped")
doctest* doctest::skip() decorator: TEST_CASE("name" * doctest::skip(true))
Boost.Testboost::unit_test::disabled() decorator, or BOOST_AUTO_TEST_CASE(name, *boost::unit_test::disabled())

DISABLED_ prefix without a tracking comment is a smell — flag as Ignored Test.

Exception Handling — Idiomatic Alternatives

// GoogleTest:
EXPECT_THROW({
    service.placeOrder(empty);
}, InvalidOrderException);

// Or capture and inspect:
try {
    service.placeOrder(empty);
    FAIL() << "Expected InvalidOrderException";
} catch (const InvalidOrderException& e) {
    EXPECT_STREQ("at least one item", e.what());
}

// Catch2:
REQUIRE_THROWS_AS(service.placeOrder(empty), InvalidOrderException);
REQUIRE_THROWS_WITH(service.placeOrder(empty), Catch::Contains("at least one item"));

The manual try/catch/FAIL pattern is acceptable when message inspection is needed; flag bare try { ... } catch (...) {} (swallowed).

Mystery Guest — Common C++ Patterns

IndicatorWhat to look for
File systemstd::ifstream, std::ofstream, fopen, hard-coded paths
Networkraw socket() / connect(), curl_easy_perform to real URL
Environmentstd::getenv("X"), Windows registry calls
Databasedirect sqlite3_open(path), ODBC connections
Acceptablestd::stringstream, std::tmpfile, GoogleMock for collaborators, boost::iostreams, in-memory streams

Integration Test Markers

  • File suffix: *_integration_test.cc, *_e2e_test.cpp

  • GoogleTest suite names containing Integration / EndToEnd

  • Catch2 tags: [integration], [e2e], [slow]

  • CMake target names ending in _integration_tests

  • Conditional compilation: #ifdef BUILD_INTEGRATION_TESTS

Setup/Teardown

FrameworkPer-testPer-suite
GoogleTest fixturevoid SetUp() overridestatic void SetUpTestSuite()
GoogleTest fixturevoid TearDown() overridestatic void TearDownTestSuite()
Catch2TEST_CASE body + SECTION re-runs setup per sectionfixture class via TEST_CASE_METHOD(Fixture, "name")
doctestsimilar to Catch2doctest::TestCase fixture
Boost.TestBOOST_FIXTURE_TEST_CASE(name, Fixture)BOOST_GLOBAL_FIXTURE(Fixture)

Catch2 SECTIONs are re-entered for each combination, so the TEST_CASE body acts as fresh per-section setup — a powerful idiom.

Tag/Trait Attributes (for test-tagging)

FrameworkTag mechanismExample
Catch2[tag] syntax in TEST_CASE second argTEST_CASE("creates order", "[positive][critical-path]")
doctest* doctest::test_suite("tag") decorator chainTEST_CASE("name" * doctest::test_suite("positive"))
GoogleTesttest name prefix convention (e.g., Positive_*, Boundary_*) or --gtest_filter patternssuite naming or TEST(PositiveCases, ...); report-only for auto-edit
Boost.Testlabel decorator: * boost::unit_test::label("positive")BOOST_AUTO_TEST_CASE(name, *boost::unit_test::label("positive"))

Filter syntax:

  • Catch2: ./tests "[positive]" ~"[slow]"

  • doctest: ./tests -ts="positive"

  • GoogleTest: ./tests --gtest_filter='Positive*'

  • Boost.Test: ./tests --run_test=@positive

Language-specific calibration notes

  • EXPECT_* continues on failure in GoogleTest — many EXPECT_EQ calls in one test may produce cascading messages from one root cause.

  • REQUIRE_* / ASSERT_* aborts — use for preconditions in long tests.

  • Death tests (EXPECT_DEATH) fork the process and check stderr — slow; acknowledge as integration-style.

  • DISABLED_ prefix disables tests silently — --gtest_also_run_disabled_tests is required to opt back in. Flag committed DISABLED_ tests as Ignored Test.

  • Catch2 SECTIONs are NOT duplicate tests — each section is a permutation of the parent TEST_CASE.

  • GoogleMock EXPECT_CALL(mock, Method(...)) counts as a state/side-effect assertion.

  • Template / typed tests (TYPED_TEST, TEMPLATE_TEST_CASE) are parametrized, not duplicates.

  • Hidden tests (Catch2 [.] or [!hide]) are excluded by default but runnable on demand — note in audit.

  • Sanitizer-only tests (#ifdef __SANITIZE_THREAD__, etc.) are conditional smoke checks — note but don't flag.

  • Test binaries that don't link gtest_main require a custom main() — verify it calls RUN_ALL_TESTS().

  • SUCCEED() / INFO(...) are not assertions; tests with only SUCCEED() are assertion-free.

Requirements and compatibility

Written for: Claude Code

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Needs: Claude Code frontmatter, Needs Python, Needs Node.js

Security

No issues found by our static scan· No issues found

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Attribution and licence

Published by
.NET (Microsoft)
Publisher
.NET (Microsoft) (Official)
Source
dotnet/skills · plugins/dotnet-test/skills/test-analysis-extensions · commit 96c3d68
Licence
MIT License (redistribution allowed)
  • LICENSE (added by AutoGineer from LICENSE)
  • ATTRIBUTION.md (written by AutoGineer)
Synced

AutoGineer is not affiliated with or endorsed by .NET (Microsoft) or .NET (Microsoft).

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download test-analysis-extensions.zip