[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M786ZDVAWD0REPHAFDT68G89":3,"skill-same:01M786ZDVAWD0REPHAFDT68G89":46},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":26,"commitSha":27,"skillDirectory":28,"forksCount":14,"repositoryHtmlUrl":29,"repositoryHomepageUrl":6,"ownerHtmlUrl":30,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":31,"install":32,"resources":45},"01M786ZDVAWD0REPHAFDT68G89","cpp-testing",null,"Use ONLY when user explicitly requests C++ test code, GoogleTest\u002FCatch2\u002FCTest configuration,\ntest failure diagnosis, or coverage\u002Fsanitizer for C++ projects.\n\nTriggers: \"write test cpp\", \"gtest\", \"GoogleTest\", \"catch2\", \"doctest\",\n\"C++ coverage\", \"sanitizer\", \"ubsan\", \"asan\", \"C++ mock\".\n\nDo NOT activate for: general C++ coding without test context, CMake build\nconfiguration, C++ template\u002Fcompile errors, or header-only library setup.","chenziyang110\u002Flaunchdeck","chenziyang110","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F43693913?v=4","01M7863EBWK1D0P6A29YGEMAS4","Robin11","01M7863ECM1VB6RXDMHTW90D2T",0,"MIT","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fcpp-testing\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: cpp-testing\ndescription: |\n  Use ONLY when user explicitly requests C++ test code, GoogleTest\u002FCatch2\u002FCTest configuration,\n  test failure diagnosis, or coverage\u002Fsanitizer for C++ projects.\n\n  Triggers: \"write test cpp\", \"gtest\", \"GoogleTest\", \"catch2\", \"doctest\",\n  \"C++ coverage\", \"sanitizer\", \"ubsan\", \"asan\", \"C++ mock\".\n\n  Do NOT activate for: general C++ coding without test context, CMake build\n  configuration, C++ template\u002Fcompile errors, or header-only library setup.\norigin: ECC\n---\n\n# C++ Testing (Agent Skill)\n\nAgent-focused testing workflow for modern C++ (C++17\u002F20) using GoogleTest\u002FGoogleMock with CMake\u002FCTest.\n\n## When to Use\n\n- Writing new C++ tests or fixing existing tests\n- Designing unit\u002Fintegration test coverage for C++ components\n- Adding test coverage, CI gating, or regression protection\n- Configuring CMake\u002FCTest workflows for consistent execution\n- Investigating test failures or flaky behavior\n- Enabling sanitizers for memory\u002Frace diagnostics\n\n### When NOT to Use\n\n- Implementing new product features without test changes\n- Large-scale refactors unrelated to test coverage or failures\n- Performance tuning without test regressions to validate\n- Non-C++ projects or non-test tasks\n\n### Stop Conditions — When to ASK instead of generating\n\nBefore generating any test code, STOP and ask the user if:\n- The code under test has **zero testability** (global state, hard-coded I\u002FO, no virtual\u002Finterface boundaries)\n- **No test framework** is detected in CMakeLists.txt (no GTest\u002FCatch2\u002FCTest)\n- The request is **ambiguous**: \"test my code\" or \"add tests\" without specifying files or classes\n- A **required test dependency** (GTest, Catch2) is not declared in CMake\n- The test would require **secrets, real credentials, or production database access**\n\n## Core Concepts\n\n- **TDD loop**: red → green → refactor (tests first, minimal fix, then cleanups).\n- **Isolation**: prefer dependency injection and fakes over global state.\n- **Test layout**: `tests\u002Funit`, `tests\u002Fintegration`, `tests\u002Ftestdata`.\n- **Mocks vs fakes**: mock for interactions, fake for stateful behavior.\n- **CTest discovery**: use `gtest_discover_tests()` for stable test discovery.\n- **CI signal**: run subset first, then full suite with `--output-on-failure`.\n\n## Agent Execution Protocol\n\nWhen this skill is activated, follow this exact execution sequence:\n\n### Phase 1: Context Analysis\nBefore writing any test code:\n1. Identify the C++ standard (17\u002F20\u002F23) and build tool (CMake preferred).\n2. Check for existing test files (`tests\u002F`, `*_test.cpp`) to infer naming conventions and style.\n3. Analyze the code under test: identify public API surface, side effects, memory management patterns, and concurrency.\n4. Determine request type: (a) new tests, (b) fixing broken tests, (c) coverage\u002Fsanitizers, (d) flaky test diagnosis.\n\n### Phase 2: Strategy Selection\nBased on Phase 1, select the appropriate strategy:\n- **Pure functions \u002F value objects**: unit tests only, no mocks needed. Use `TEST`\u002F`TEST_F`.\n- **External dependencies (I\u002FO, network)**: use GoogleMock (`EXPECT_CALL`) or fakes for stateful behavior.\n- **Memory-critical code**: enable AddressSanitizer (ASan) and UndefinedBehaviorSanitizer (UBSan) in CI.\n- **Concurrent code**: use ThreadSanitizer (TSan); avoid sleeps, use condition variables\u002Flatches.\n\n### Decision Tree (Complex Scenarios)\n\n```text\nQ1: External I\u002FO (filesystem, network, hardware)?\n  → Yes → Q2\n  → No  → Pure unit test, no mocks needed\n\nQ2: Mockable (virtual interface, template parameter, function pointer)?\n  → Yes → Mock at boundary (GoogleMock EXPECT_CALL, MOCK_METHOD)\n  → No  → Q3\n\nQ3: Needs real behavior (memory, concurrency)?\n  → Yes → Integration test + ASan\u002FUBSan\u002FTSan in CI\n  → No  → Use fake\u002Ftest double\n\nQ4: Memory-critical \u002F UB risk?\n  → Yes → Enable AddressSanitizer + UndefinedBehaviorSanitizer\n  → No  → Standard TEST\u002FTEST_F\n\nQ5: Concurrency test?\n  → Yes → ThreadSanitizer + deterministic thread scheduling\n  → No  → Standard unit test\n```\n\n### Phase 3: Test Case Design\nApply systematic test design before coding:\n1. **Happy path**: normal input → expected output.\n2. **Boundary values**: empty containers, `nullptr`, `std::nullopt`, max container size, numeric limits.\n3. **Error cases**: exceptions, error codes, invalid arguments, resource exhaustion.\n4. **Memory & concurrency**: memory leaks (use `testing::internal::Leakable` pattern), data races, RAII correctness.\n\n### Phase 4: Code Generation\nOutput tests in this exact structure:\n```\n#### Test Strategy Summary\n- [ ] Scope: Unit \u002F Integration\n- [ ] Framework: GoogleTest + GoogleMock + CMake\u002FCTest\n- [ ] Mocking: Yes\u002FNo — GoogleMock \u002F fakes\n- [ ] Edge cases covered: \u003Clist>\n\n#### Generated Test Code\n\u003Cfile_path>\n\u003Ccode>\n\n#### Running Commands\n```bash\nctest --test-dir build --output-on-failure\n.\u002Fbuild\u002Fexample_tests --gtest_filter=TestSuite.TestName\n```\n\n#### Coverage Impact (if applicable)\n- Estimated coverage change: \u003CX%>\n- Command to verify: `cmake -S . -B build-cov -DENABLE_COVERAGE=ON && cmake --build build-cov && ctest --test-dir build-cov`\n```\n\n### Phase 5: Verification Checklist\nBefore completing, verify:\n- [ ] Tests compile without errors (`cmake --build build`).\n- [ ] Each test has exactly one logical assertion focus (use `ASSERT_*` for preconditions, `EXPECT_*` for checks).\n- [ ] Mocks are reset between tests (reconstruct fixtures in `SetUp`).\n- [ ] No `sleep` used for synchronization (use condition variables or latches).\n- [ ] Flaky test guardrails applied (unique temp dirs, deterministic seeds, ASan\u002FUBSan in CI).\n\n## TDD Workflow\n\nFollow the RED → GREEN → REFACTOR loop:\n\n1. **RED**: write a failing test that captures the new behavior\n2. **GREEN**: implement the smallest change to pass\n3. **REFACTOR**: clean up while tests stay green\n\n```cpp\n\u002F\u002F tests\u002Fadd_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nint Add(int a, int b); \u002F\u002F Provided by production code.\n\nTEST(AddTest, AddsTwoNumbers) { \u002F\u002F RED\n  EXPECT_EQ(Add(2, 3), 5);\n}\n\n\u002F\u002F src\u002Fadd.cpp\nint Add(int a, int b) { \u002F\u002F GREEN\n  return a + b;\n}\n\n\u002F\u002F REFACTOR: simplify\u002Frename once tests pass\n```\n\n## Code Examples\n\n### Basic Unit Test (gtest)\n\n```cpp\n\u002F\u002F tests\u002Fcalculator_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nint Add(int a, int b); \u002F\u002F Provided by production code.\n\nTEST(CalculatorTest, AddsTwoNumbers) {\n    EXPECT_EQ(Add(2, 3), 5);\n}\n```\n\n### Fixture (gtest)\n\n```cpp\n\u002F\u002F tests\u002Fuser_store_test.cpp\n\u002F\u002F Pseudocode stub: replace UserStore\u002FUser with project types.\n#include \u003Cgtest\u002Fgtest.h>\n#include \u003Cmemory>\n#include \u003Coptional>\n#include \u003Cstring>\n\nstruct User { std::string name; };\nclass UserStore {\npublic:\n    explicit UserStore(std::string \u002F*path*\u002F) {}\n    void Seed(std::initializer_list\u003CUser> \u002F*users*\u002F) {}\n    std::optional\u003CUser> Find(const std::string &\u002F*name*\u002F) { return User{\"alice\"}; }\n};\n\nclass UserStoreTest : public ::testing::Test {\nprotected:\n    void SetUp() override {\n        store = std::make_unique\u003CUserStore>(\":memory:\");\n        store->Seed({{\"alice\"}, {\"bob\"}});\n    }\n\n    std::unique_ptr\u003CUserStore> store;\n};\n\nTEST_F(UserStoreTest, FindsExistingUser) {\n    auto user = store->Find(\"alice\");\n    ASSERT_TRUE(user.has_value());\n    EXPECT_EQ(user->name, \"alice\");\n}\n```\n\n### Mock (gmock)\n\n```cpp\n\u002F\u002F tests\u002Fnotifier_test.cpp\n#include \u003Cgmock\u002Fgmock.h>\n#include \u003Cgtest\u002Fgtest.h>\n#include \u003Cstring>\n\nclass Notifier {\npublic:\n    virtual ~Notifier() = default;\n    virtual void Send(const std::string &message) = 0;\n};\n\nclass MockNotifier : public Notifier {\npublic:\n    MOCK_METHOD(void, Send, (const std::string &message), (override));\n};\n\nclass Service {\npublic:\n    explicit Service(Notifier &notifier) : notifier_(notifier) {}\n    void Publish(const std::string &message) { notifier_.Send(message); }\n\nprivate:\n    Notifier &notifier_;\n};\n\nTEST(ServiceTest, SendsNotifications) {\n    MockNotifier notifier;\n    Service service(notifier);\n\n    EXPECT_CALL(notifier, Send(\"hello\")).Times(1);\n    service.Publish(\"hello\");\n}\n```\n\n### CMake\u002FCTest Quickstart\n\n```cmake\n# CMakeLists.txt (excerpt)\ncmake_minimum_required(VERSION 3.20)\nproject(example LANGUAGES CXX)\n\nset(CMAKE_CXX_STANDARD 20)\nset(CMAKE_CXX_STANDARD_REQUIRED ON)\n\ninclude(FetchContent)\n# Prefer project-locked versions. If using a tag, use a pinned version per project policy.\nset(GTEST_VERSION v1.15.2) # Latest stable as of 2025; verify at github.com\u002Fgoogle\u002Fgoogletest\nFetchContent_Declare(\n  googletest\n  # Google Test framework (official repository)\n  URL https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fgoogletest\u002Farchive\u002Frefs\u002Ftags\u002F${GTEST_VERSION}.zip\n)\nFetchContent_MakeAvailable(googletest)\n\nadd_executable(example_tests\n  tests\u002Fcalculator_test.cpp\n  src\u002Fcalculator.cpp\n)\ntarget_link_libraries(example_tests GTest::gtest GTest::gmock GTest::gtest_main)\n\nenable_testing()\ninclude(GoogleTest)\ngtest_discover_tests(example_tests)\n```\n\n```bash\ncmake -S . -B build -DCMAKE_BUILD_TYPE=Debug\ncmake --build build -j\nctest --test-dir build --output-on-failure\n```\n\n## Running Tests\n\n```bash\nctest --test-dir build --output-on-failure\nctest --test-dir build -R ClampTest\nctest --test-dir build -R \"UserStoreTest.*\" --output-on-failure\n```\n\n```bash\n.\u002Fbuild\u002Fexample_tests --gtest_filter=ClampTest.*\n.\u002Fbuild\u002Fexample_tests --gtest_filter=UserStoreTest.FindsExistingUser\n```\n\n## Debugging Failures\n\n1. Re-run the single failing test with gtest filter.\n2. Add scoped logging around the failing assertion.\n3. Re-run with sanitizers enabled.\n4. Expand to full suite once the root cause is fixed.\n\n## Coverage\n\nPrefer target-level settings instead of global flags.\n\n```cmake\noption(ENABLE_COVERAGE \"Enable coverage flags\" OFF)\n\nif(ENABLE_COVERAGE)\n  if(CMAKE_CXX_COMPILER_ID MATCHES \"GNU\")\n    target_compile_options(example_tests PRIVATE --coverage)\n    target_link_options(example_tests PRIVATE --coverage)\n  elseif(CMAKE_CXX_COMPILER_ID MATCHES \"Clang\")\n    target_compile_options(example_tests PRIVATE -fprofile-instr-generate -fcoverage-mapping)\n    target_link_options(example_tests PRIVATE -fprofile-instr-generate)\n  endif()\nendif()\n```\n\nGCC + gcov + lcov:\n\n```bash\ncmake -S . -B build-cov -DENABLE_COVERAGE=ON\ncmake --build build-cov -j\nctest --test-dir build-cov\nlcov --capture --directory build-cov --output-file coverage.info\nlcov --remove coverage.info '\u002Fusr\u002F*' --output-file coverage.info\ngenhtml coverage.info --output-directory coverage\n```\n\nClang + llvm-cov:\n\n```bash\ncmake -S . -B build-llvm -DENABLE_COVERAGE=ON -DCMAKE_CXX_COMPILER=clang++\ncmake --build build-llvm -j\nLLVM_PROFILE_FILE=\"build-llvm\u002Fdefault.profraw\" ctest --test-dir build-llvm\nllvm-profdata merge -sparse build-llvm\u002Fdefault.profraw -o build-llvm\u002Fdefault.profdata\nllvm-cov report build-llvm\u002Fexample_tests -instr-profile=build-llvm\u002Fdefault.profdata\n```\n\n## Sanitizers\n\n```cmake\noption(ENABLE_ASAN \"Enable AddressSanitizer\" OFF)\noption(ENABLE_UBSAN \"Enable UndefinedBehaviorSanitizer\" OFF)\noption(ENABLE_TSAN \"Enable ThreadSanitizer\" OFF)\n\nif(ENABLE_ASAN)\n  add_compile_options(-fsanitize=address -fno-omit-frame-pointer)\n  add_link_options(-fsanitize=address)\nendif()\nif(ENABLE_UBSAN)\n  add_compile_options(-fsanitize=undefined -fno-omit-frame-pointer)\n  add_link_options(-fsanitize=undefined)\nendif()\nif(ENABLE_TSAN)\n  add_compile_options(-fsanitize=thread)\n  add_link_options(-fsanitize=thread)\nendif()\n```\n\n## CI\u002FCD Integration (GitHub Actions)\n\n```yaml\n# .github\u002Fworkflows\u002Ftest.yml\nname: Test\non: [push, pull_request]\njobs:\n  test:\n    runs-on: ubuntu-latest\n    steps:\n      - uses: actions\u002Fcheckout@v4\n      - run: cmake -B build -DCMAKE_BUILD_TYPE=Debug\n      - run: cmake --build build\n      - run: cd build && ctest --output-on-failure\n```\n\n## Flaky Tests Guardrails\n\n- Never use `sleep` for synchronization; use condition variables or latches.\n- Make temp directories unique per test and always clean them.\n- Avoid real time, network, or filesystem dependencies in unit tests.\n- Use deterministic seeds for randomized inputs.\n\n## Test Case Design Patterns\n\nApply these systematic patterns before writing test code to ensure robust coverage beyond the happy path.\n\n### 1. Boundary Value Analysis\nTest at the edges of valid input domains: empty, zero, maximum, null, and just-outside valid ranges.\n\n```cpp\n\u002F\u002F tests\u002Fvalidator_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n#include \u003Cstring>\n\nTEST(ValidatorTest, RejectsInvalidInput) {\n    EXPECT_THROW(validator.validate(\"\"), std::invalid_argument);\n    EXPECT_THROW(validator.validate(\" \"), std::invalid_argument);\n    EXPECT_THROW(validator.validate(std::string(10001, 'a')), std::invalid_argument);\n}\n```\n\n### 2. Negative \u002F Error Case Testing\nVerify the system behaves correctly under invalid or unexpected conditions.\n\n```cpp\n\u002F\u002F tests\u002Fcalculator_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nTEST(CalculatorTest, ThrowsOnDivideByZero) {\n    EXPECT_THROW(calculator.divide(10, 0), std::runtime_error);\n}\n```\n\n### 3. State-Based Testing\nFor objects with internal state, verify transitions between states are correct and irreversible where expected.\n\n```cpp\n\u002F\u002F tests\u002Fconnection_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nTEST(ConnectionTest, StateTransitionsAreCorrect) {\n    Connection conn;\n    EXPECT_EQ(conn.state(), State::CLOSED);\n    conn.open();\n    EXPECT_EQ(conn.state(), State::OPEN);\n    conn.close();\n    EXPECT_EQ(conn.state(), State::CLOSED);\n    \u002F\u002F Idempotency\n    conn.close();\n    EXPECT_EQ(conn.state(), State::CLOSED);\n}\n```\n\n## Common Test Smells to Avoid\n\n- **Assertion Roulette**: Many assertions without descriptive messages, making failure diagnosis ambiguous. *Fix: Use `EXPECT_EQ(val1, val2) \u003C\u003C \"context message\"` or `ASSERT_*` for critical checks.*\n- **Happy Path Bias**: Only testing successful scenarios, leaving error paths uncovered. *Fix: Test invalid inputs, edge values, `EXPECT_THROW` for exceptions.*\n- **Mock Overdose**: Too many mocks hiding integration issues or coupling tests to implementation. *Fix: Mock only at interface\u002Fvirtual boundaries; prefer fakes for stateful behavior.*\n- **The Free Ride**: Tests that execute code but assert nothing meaningful. *Fix: Every test must have at least one specific, falsifiable assertion.*\n- **Obscure Test**: Long test cases with no clear fixture\u002Ftest split. *Fix: Use `TEST_F` for shared setup, keep individual `TEST` methods short.*\n- **Flaky Test**: Non-deterministic results from memory, threads, or unsanitized UB. *Fix: Enable ASan\u002FUBSan\u002FTSan in CI; use deterministic inputs; avoid shared mutable state.*\n\n## Best Practices\n\n### DO\n\n- Keep tests deterministic and isolated\n- Prefer dependency injection over globals\n- Use `ASSERT_*` for preconditions, `EXPECT_*` for multiple checks\n- Separate unit vs integration tests in CTest labels or directories\n- Run sanitizers in CI for memory and race detection\n\n### DON'T\n\n- Don't depend on real time or network in unit tests\n- Don't use sleeps as synchronization when a condition variable can be used\n- Don't over-mock simple value objects\n- Don't use brittle string matching for non-critical logs\n\n### Common Pitfalls\n\n- **Using fixed temp paths** → Generate unique temp directories per test and clean them.\n- **Relying on wall clock time** → Inject a clock or use fake time sources.\n- **Flaky concurrency tests** → Use condition variables\u002Flatches and bounded waits.\n- **Hidden global state** → Reset global state in fixtures or remove globals.\n- **Over-mocking** → Prefer fakes for stateful behavior and only mock interactions.\n- **Missing sanitizer runs** → Add ASan\u002FUBSan\u002FTSan builds in CI.\n- **Coverage on debug-only builds** → Ensure coverage targets use consistent flags.\n\n## Optional Appendix: Fuzzing \u002F Property Testing\n\nOnly use if the project already supports LLVM\u002FlibFuzzer or a property-testing library.\n\n- **libFuzzer**: best for pure functions with minimal I\u002FO.\n- **RapidCheck**: property-based tests to validate invariants.\n\nMinimal libFuzzer harness (pseudocode: replace ParseConfig):\n\n```cpp\n#include \u003Ccstddef>\n#include \u003Ccstdint>\n#include \u003Cstring>\n\nextern \"C\" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {\n    std::string input(reinterpret_cast\u003Cconst char *>(data), size);\n    \u002F\u002F ParseConfig(input); \u002F\u002F project function\n    return 0;\n}\n```\n\n### Mutation Testing (Mull)\n\n```bash\n# Mull: LLVM-based mutation testing for C\u002FC++\n# https:\u002F\u002Fgithub.com\u002Fmull-project\u002Fmull\nmull-cxx -test-framework=GoogleTest -compilation-database-path=build\u002Fcompile_commands.json\n# Report: mutation score per function with killed\u002Fsurvived mutants\n```\n\n> Mutation testing introduces code changes (e.g., `+` → `-`, removing `const`) to verify tests detect them. **Target ≥80% mutation score** for critical code paths.\n\n## Alternatives to GoogleTest\n\n- **Catch2** (v3.7.x): header-only, expressive matchers\n- **doctest** (v2.4.x): lightweight, minimal compile overhead\n","# C++ Testing (Agent Skill)\n\nAgent-focused testing workflow for modern C++ (C++17\u002F20) using GoogleTest\u002FGoogleMock with CMake\u002FCTest.\n\n## When to Use\n\n- Writing new C++ tests or fixing existing tests\n- Designing unit\u002Fintegration test coverage for C++ components\n- Adding test coverage, CI gating, or regression protection\n- Configuring CMake\u002FCTest workflows for consistent execution\n- Investigating test failures or flaky behavior\n- Enabling sanitizers for memory\u002Frace diagnostics\n\n### When NOT to Use\n\n- Implementing new product features without test changes\n- Large-scale refactors unrelated to test coverage or failures\n- Performance tuning without test regressions to validate\n- Non-C++ projects or non-test tasks\n\n### Stop Conditions — When to ASK instead of generating\n\nBefore generating any test code, STOP and ask the user if:\n- The code under test has **zero testability** (global state, hard-coded I\u002FO, no virtual\u002Finterface boundaries)\n- **No test framework** is detected in CMakeLists.txt (no GTest\u002FCatch2\u002FCTest)\n- The request is **ambiguous**: \"test my code\" or \"add tests\" without specifying files or classes\n- A **required test dependency** (GTest, Catch2) is not declared in CMake\n- The test would require **secrets, real credentials, or production database access**\n\n## Core Concepts\n\n- **TDD loop**: red → green → refactor (tests first, minimal fix, then cleanups).\n- **Isolation**: prefer dependency injection and fakes over global state.\n- **Test layout**: `tests\u002Funit`, `tests\u002Fintegration`, `tests\u002Ftestdata`.\n- **Mocks vs fakes**: mock for interactions, fake for stateful behavior.\n- **CTest discovery**: use `gtest_discover_tests()` for stable test discovery.\n- **CI signal**: run subset first, then full suite with `--output-on-failure`.\n\n## Agent Execution Protocol\n\nWhen this skill is activated, follow this exact execution sequence:\n\n### Phase 1: Context Analysis\nBefore writing any test code:\n1. Identify the C++ standard (17\u002F20\u002F23) and build tool (CMake preferred).\n2. Check for existing test files (`tests\u002F`, `*_test.cpp`) to infer naming conventions and style.\n3. Analyze the code under test: identify public API surface, side effects, memory management patterns, and concurrency.\n4. Determine request type: (a) new tests, (b) fixing broken tests, (c) coverage\u002Fsanitizers, (d) flaky test diagnosis.\n\n### Phase 2: Strategy Selection\nBased on Phase 1, select the appropriate strategy:\n- **Pure functions \u002F value objects**: unit tests only, no mocks needed. Use `TEST`\u002F`TEST_F`.\n- **External dependencies (I\u002FO, network)**: use GoogleMock (`EXPECT_CALL`) or fakes for stateful behavior.\n- **Memory-critical code**: enable AddressSanitizer (ASan) and UndefinedBehaviorSanitizer (UBSan) in CI.\n- **Concurrent code**: use ThreadSanitizer (TSan); avoid sleeps, use condition variables\u002Flatches.\n\n### Decision Tree (Complex Scenarios)\n\n```text\nQ1: External I\u002FO (filesystem, network, hardware)?\n  → Yes → Q2\n  → No  → Pure unit test, no mocks needed\n\nQ2: Mockable (virtual interface, template parameter, function pointer)?\n  → Yes → Mock at boundary (GoogleMock EXPECT_CALL, MOCK_METHOD)\n  → No  → Q3\n\nQ3: Needs real behavior (memory, concurrency)?\n  → Yes → Integration test + ASan\u002FUBSan\u002FTSan in CI\n  → No  → Use fake\u002Ftest double\n\nQ4: Memory-critical \u002F UB risk?\n  → Yes → Enable AddressSanitizer + UndefinedBehaviorSanitizer\n  → No  → Standard TEST\u002FTEST_F\n\nQ5: Concurrency test?\n  → Yes → ThreadSanitizer + deterministic thread scheduling\n  → No  → Standard unit test\n```\n\n### Phase 3: Test Case Design\nApply systematic test design before coding:\n1. **Happy path**: normal input → expected output.\n2. **Boundary values**: empty containers, `nullptr`, `std::nullopt`, max container size, numeric limits.\n3. **Error cases**: exceptions, error codes, invalid arguments, resource exhaustion.\n4. **Memory & concurrency**: memory leaks (use `testing::internal::Leakable` pattern), data races, RAII correctness.\n\n### Phase 4: Code Generation\nOutput tests in this exact structure:\n```\n#### Test Strategy Summary\n- [ ] Scope: Unit \u002F Integration\n- [ ] Framework: GoogleTest + GoogleMock + CMake\u002FCTest\n- [ ] Mocking: Yes\u002FNo — GoogleMock \u002F fakes\n- [ ] Edge cases covered: \u003Clist>\n\n#### Generated Test Code\n\u003Cfile_path>\n\u003Ccode>\n\n#### Running Commands\n```bash\nctest --test-dir build --output-on-failure\n.\u002Fbuild\u002Fexample_tests --gtest_filter=TestSuite.TestName\n```\n\n#### Coverage Impact (if applicable)\n- Estimated coverage change: \u003CX%>\n- Command to verify: `cmake -S . -B build-cov -DENABLE_COVERAGE=ON && cmake --build build-cov && ctest --test-dir build-cov`\n```\n\n### Phase 5: Verification Checklist\nBefore completing, verify:\n- [ ] Tests compile without errors (`cmake --build build`).\n- [ ] Each test has exactly one logical assertion focus (use `ASSERT_*` for preconditions, `EXPECT_*` for checks).\n- [ ] Mocks are reset between tests (reconstruct fixtures in `SetUp`).\n- [ ] No `sleep` used for synchronization (use condition variables or latches).\n- [ ] Flaky test guardrails applied (unique temp dirs, deterministic seeds, ASan\u002FUBSan in CI).\n\n## TDD Workflow\n\nFollow the RED → GREEN → REFACTOR loop:\n\n1. **RED**: write a failing test that captures the new behavior\n2. **GREEN**: implement the smallest change to pass\n3. **REFACTOR**: clean up while tests stay green\n\n```cpp\n\u002F\u002F tests\u002Fadd_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nint Add(int a, int b); \u002F\u002F Provided by production code.\n\nTEST(AddTest, AddsTwoNumbers) { \u002F\u002F RED\n  EXPECT_EQ(Add(2, 3), 5);\n}\n\n\u002F\u002F src\u002Fadd.cpp\nint Add(int a, int b) { \u002F\u002F GREEN\n  return a + b;\n}\n\n\u002F\u002F REFACTOR: simplify\u002Frename once tests pass\n```\n\n## Code Examples\n\n### Basic Unit Test (gtest)\n\n```cpp\n\u002F\u002F tests\u002Fcalculator_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nint Add(int a, int b); \u002F\u002F Provided by production code.\n\nTEST(CalculatorTest, AddsTwoNumbers) {\n    EXPECT_EQ(Add(2, 3), 5);\n}\n```\n\n### Fixture (gtest)\n\n```cpp\n\u002F\u002F tests\u002Fuser_store_test.cpp\n\u002F\u002F Pseudocode stub: replace UserStore\u002FUser with project types.\n#include \u003Cgtest\u002Fgtest.h>\n#include \u003Cmemory>\n#include \u003Coptional>\n#include \u003Cstring>\n\nstruct User { std::string name; };\nclass UserStore {\npublic:\n    explicit UserStore(std::string \u002F*path*\u002F) {}\n    void Seed(std::initializer_list\u003CUser> \u002F*users*\u002F) {}\n    std::optional\u003CUser> Find(const std::string &\u002F*name*\u002F) { return User{\"alice\"}; }\n};\n\nclass UserStoreTest : public ::testing::Test {\nprotected:\n    void SetUp() override {\n        store = std::make_unique\u003CUserStore>(\":memory:\");\n        store->Seed({{\"alice\"}, {\"bob\"}});\n    }\n\n    std::unique_ptr\u003CUserStore> store;\n};\n\nTEST_F(UserStoreTest, FindsExistingUser) {\n    auto user = store->Find(\"alice\");\n    ASSERT_TRUE(user.has_value());\n    EXPECT_EQ(user->name, \"alice\");\n}\n```\n\n### Mock (gmock)\n\n```cpp\n\u002F\u002F tests\u002Fnotifier_test.cpp\n#include \u003Cgmock\u002Fgmock.h>\n#include \u003Cgtest\u002Fgtest.h>\n#include \u003Cstring>\n\nclass Notifier {\npublic:\n    virtual ~Notifier() = default;\n    virtual void Send(const std::string &message) = 0;\n};\n\nclass MockNotifier : public Notifier {\npublic:\n    MOCK_METHOD(void, Send, (const std::string &message), (override));\n};\n\nclass Service {\npublic:\n    explicit Service(Notifier &notifier) : notifier_(notifier) {}\n    void Publish(const std::string &message) { notifier_.Send(message); }\n\nprivate:\n    Notifier &notifier_;\n};\n\nTEST(ServiceTest, SendsNotifications) {\n    MockNotifier notifier;\n    Service service(notifier);\n\n    EXPECT_CALL(notifier, Send(\"hello\")).Times(1);\n    service.Publish(\"hello\");\n}\n```\n\n### CMake\u002FCTest Quickstart\n\n```cmake\n# CMakeLists.txt (excerpt)\ncmake_minimum_required(VERSION 3.20)\nproject(example LANGUAGES CXX)\n\nset(CMAKE_CXX_STANDARD 20)\nset(CMAKE_CXX_STANDARD_REQUIRED ON)\n\ninclude(FetchContent)\n# Prefer project-locked versions. If using a tag, use a pinned version per project policy.\nset(GTEST_VERSION v1.15.2) # Latest stable as of 2025; verify at github.com\u002Fgoogle\u002Fgoogletest\nFetchContent_Declare(\n  googletest\n  # Google Test framework (official repository)\n  URL https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fgoogletest\u002Farchive\u002Frefs\u002Ftags\u002F${GTEST_VERSION}.zip\n)\nFetchContent_MakeAvailable(googletest)\n\nadd_executable(example_tests\n  tests\u002Fcalculator_test.cpp\n  src\u002Fcalculator.cpp\n)\ntarget_link_libraries(example_tests GTest::gtest GTest::gmock GTest::gtest_main)\n\nenable_testing()\ninclude(GoogleTest)\ngtest_discover_tests(example_tests)\n```\n\n```bash\ncmake -S . -B build -DCMAKE_BUILD_TYPE=Debug\ncmake --build build -j\nctest --test-dir build --output-on-failure\n```\n\n## Running Tests\n\n```bash\nctest --test-dir build --output-on-failure\nctest --test-dir build -R ClampTest\nctest --test-dir build -R \"UserStoreTest.*\" --output-on-failure\n```\n\n```bash\n.\u002Fbuild\u002Fexample_tests --gtest_filter=ClampTest.*\n.\u002Fbuild\u002Fexample_tests --gtest_filter=UserStoreTest.FindsExistingUser\n```\n\n## Debugging Failures\n\n1. Re-run the single failing test with gtest filter.\n2. Add scoped logging around the failing assertion.\n3. Re-run with sanitizers enabled.\n4. Expand to full suite once the root cause is fixed.\n\n## Coverage\n\nPrefer target-level settings instead of global flags.\n\n```cmake\noption(ENABLE_COVERAGE \"Enable coverage flags\" OFF)\n\nif(ENABLE_COVERAGE)\n  if(CMAKE_CXX_COMPILER_ID MATCHES \"GNU\")\n    target_compile_options(example_tests PRIVATE --coverage)\n    target_link_options(example_tests PRIVATE --coverage)\n  elseif(CMAKE_CXX_COMPILER_ID MATCHES \"Clang\")\n    target_compile_options(example_tests PRIVATE -fprofile-instr-generate -fcoverage-mapping)\n    target_link_options(example_tests PRIVATE -fprofile-instr-generate)\n  endif()\nendif()\n```\n\nGCC + gcov + lcov:\n\n```bash\ncmake -S . -B build-cov -DENABLE_COVERAGE=ON\ncmake --build build-cov -j\nctest --test-dir build-cov\nlcov --capture --directory build-cov --output-file coverage.info\nlcov --remove coverage.info '\u002Fusr\u002F*' --output-file coverage.info\ngenhtml coverage.info --output-directory coverage\n```\n\nClang + llvm-cov:\n\n```bash\ncmake -S . -B build-llvm -DENABLE_COVERAGE=ON -DCMAKE_CXX_COMPILER=clang++\ncmake --build build-llvm -j\nLLVM_PROFILE_FILE=\"build-llvm\u002Fdefault.profraw\" ctest --test-dir build-llvm\nllvm-profdata merge -sparse build-llvm\u002Fdefault.profraw -o build-llvm\u002Fdefault.profdata\nllvm-cov report build-llvm\u002Fexample_tests -instr-profile=build-llvm\u002Fdefault.profdata\n```\n\n## Sanitizers\n\n```cmake\noption(ENABLE_ASAN \"Enable AddressSanitizer\" OFF)\noption(ENABLE_UBSAN \"Enable UndefinedBehaviorSanitizer\" OFF)\noption(ENABLE_TSAN \"Enable ThreadSanitizer\" OFF)\n\nif(ENABLE_ASAN)\n  add_compile_options(-fsanitize=address -fno-omit-frame-pointer)\n  add_link_options(-fsanitize=address)\nendif()\nif(ENABLE_UBSAN)\n  add_compile_options(-fsanitize=undefined -fno-omit-frame-pointer)\n  add_link_options(-fsanitize=undefined)\nendif()\nif(ENABLE_TSAN)\n  add_compile_options(-fsanitize=thread)\n  add_link_options(-fsanitize=thread)\nendif()\n```\n\n## CI\u002FCD Integration (GitHub Actions)\n\n```yaml\n# .github\u002Fworkflows\u002Ftest.yml\nname: Test\non: [push, pull_request]\njobs:\n  test:\n    runs-on: ubuntu-latest\n    steps:\n      - uses: actions\u002Fcheckout@v4\n      - run: cmake -B build -DCMAKE_BUILD_TYPE=Debug\n      - run: cmake --build build\n      - run: cd build && ctest --output-on-failure\n```\n\n## Flaky Tests Guardrails\n\n- Never use `sleep` for synchronization; use condition variables or latches.\n- Make temp directories unique per test and always clean them.\n- Avoid real time, network, or filesystem dependencies in unit tests.\n- Use deterministic seeds for randomized inputs.\n\n## Test Case Design Patterns\n\nApply these systematic patterns before writing test code to ensure robust coverage beyond the happy path.\n\n### 1. Boundary Value Analysis\nTest at the edges of valid input domains: empty, zero, maximum, null, and just-outside valid ranges.\n\n```cpp\n\u002F\u002F tests\u002Fvalidator_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n#include \u003Cstring>\n\nTEST(ValidatorTest, RejectsInvalidInput) {\n    EXPECT_THROW(validator.validate(\"\"), std::invalid_argument);\n    EXPECT_THROW(validator.validate(\" \"), std::invalid_argument);\n    EXPECT_THROW(validator.validate(std::string(10001, 'a')), std::invalid_argument);\n}\n```\n\n### 2. Negative \u002F Error Case Testing\nVerify the system behaves correctly under invalid or unexpected conditions.\n\n```cpp\n\u002F\u002F tests\u002Fcalculator_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nTEST(CalculatorTest, ThrowsOnDivideByZero) {\n    EXPECT_THROW(calculator.divide(10, 0), std::runtime_error);\n}\n```\n\n### 3. State-Based Testing\nFor objects with internal state, verify transitions between states are correct and irreversible where expected.\n\n```cpp\n\u002F\u002F tests\u002Fconnection_test.cpp\n#include \u003Cgtest\u002Fgtest.h>\n\nTEST(ConnectionTest, StateTransitionsAreCorrect) {\n    Connection conn;\n    EXPECT_EQ(conn.state(), State::CLOSED);\n    conn.open();\n    EXPECT_EQ(conn.state(), State::OPEN);\n    conn.close();\n    EXPECT_EQ(conn.state(), State::CLOSED);\n    \u002F\u002F Idempotency\n    conn.close();\n    EXPECT_EQ(conn.state(), State::CLOSED);\n}\n```\n\n## Common Test Smells to Avoid\n\n- **Assertion Roulette**: Many assertions without descriptive messages, making failure diagnosis ambiguous. *Fix: Use `EXPECT_EQ(val1, val2) \u003C\u003C \"context message\"` or `ASSERT_*` for critical checks.*\n- **Happy Path Bias**: Only testing successful scenarios, leaving error paths uncovered. *Fix: Test invalid inputs, edge values, `EXPECT_THROW` for exceptions.*\n- **Mock Overdose**: Too many mocks hiding integration issues or coupling tests to implementation. *Fix: Mock only at interface\u002Fvirtual boundaries; prefer fakes for stateful behavior.*\n- **The Free Ride**: Tests that execute code but assert nothing meaningful. *Fix: Every test must have at least one specific, falsifiable assertion.*\n- **Obscure Test**: Long test cases with no clear fixture\u002Ftest split. *Fix: Use `TEST_F` for shared setup, keep individual `TEST` methods short.*\n- **Flaky Test**: Non-deterministic results from memory, threads, or unsanitized UB. *Fix: Enable ASan\u002FUBSan\u002FTSan in CI; use deterministic inputs; avoid shared mutable state.*\n\n## Best Practices\n\n### DO\n\n- Keep tests deterministic and isolated\n- Prefer dependency injection over globals\n- Use `ASSERT_*` for preconditions, `EXPECT_*` for multiple checks\n- Separate unit vs integration tests in CTest labels or directories\n- Run sanitizers in CI for memory and race detection\n\n### DON'T\n\n- Don't depend on real time or network in unit tests\n- Don't use sleeps as synchronization when a condition variable can be used\n- Don't over-mock simple value objects\n- Don't use brittle string matching for non-critical logs\n\n### Common Pitfalls\n\n- **Using fixed temp paths** → Generate unique temp directories per test and clean them.\n- **Relying on wall clock time** → Inject a clock or use fake time sources.\n- **Flaky concurrency tests** → Use condition variables\u002Flatches and bounded waits.\n- **Hidden global state** → Reset global state in fixtures or remove globals.\n- **Over-mocking** → Prefer fakes for stateful behavior and only mock interactions.\n- **Missing sanitizer runs** → Add ASan\u002FUBSan\u002FTSan builds in CI.\n- **Coverage on debug-only builds** → Ensure coverage targets use consistent flags.\n\n## Optional Appendix: Fuzzing \u002F Property Testing\n\nOnly use if the project already supports LLVM\u002FlibFuzzer or a property-testing library.\n\n- **libFuzzer**: best for pure functions with minimal I\u002FO.\n- **RapidCheck**: property-based tests to validate invariants.\n\nMinimal libFuzzer harness (pseudocode: replace ParseConfig):\n\n```cpp\n#include \u003Ccstddef>\n#include \u003Ccstdint>\n#include \u003Cstring>\n\nextern \"C\" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {\n    std::string input(reinterpret_cast\u003Cconst char *>(data), size);\n    \u002F\u002F ParseConfig(input); \u002F\u002F project function\n    return 0;\n}\n```\n\n### Mutation Testing (Mull)\n\n```bash\n# Mull: LLVM-based mutation testing for C\u002FC++\n# https:\u002F\u002Fgithub.com\u002Fmull-project\u002Fmull\nmull-cxx -test-framework=GoogleTest -compilation-database-path=build\u002Fcompile_commands.json\n# Report: mutation score per function with killed\u002Fsurvived mutants\n```\n\n> Mutation testing introduces code changes (e.g., `+` → `-`, removing `const`) to verify tests detect them. **Target ≥80% mutation score** for critical code paths.\n\n## Alternatives to GoogleTest\n\n- **Catch2** (v3.7.x): header-only, expressive matchers\n- **doctest** (v2.4.x): lightweight, minimal compile overhead\n",{"name":5,"origin":24,"description":25},"ECC","Use ONLY when user explicitly requests C++ test code, GoogleTest\u002FCatch2\u002FCTest configuration,\ntest failure diagnosis, or coverage\u002Fsanitizer for C++ projects.\n\nTriggers: \"write test cpp\", \"gtest\", \"GoogleTest\", \"catch2\", \"doctest\",\n\"C++ coverage\", \"sanitizer\", \"ubsan\", \"asan\", \"C++ mock\".\n\nDo NOT activate for: general C++ coding without test context, CMake build\nconfiguration, C++ template\u002Fcompile errors, or header-only library setup.\n","main","021f5cbe02157229460a577e9a585f2da6d78cb0",".specify\u002Ftemplates\u002Fpassive-skills\u002Fcpp-testing","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck","https:\u002F\u002Fgithub.com\u002Fchenziyang110","2026-08-07 18:10:54",{"source":8,"skillName":5,"commands":33,"prompt":43,"zipAvailable":44},[34,37,40],{"tool":35,"command":36},"npx","npx skills add chenziyang110\u002Flaunchdeck --skill cpp-testing",{"tool":38,"command":39},"bunx","bunx skills add chenziyang110\u002Flaunchdeck --skill cpp-testing",{"tool":41,"command":42},"pnpm","pnpm dlx skills add chenziyang110\u002Flaunchdeck --skill cpp-testing","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fcpp-testing\u002FSKILL.md",true,[],[47,54,60,66,73,82],{"skillNo":48,"name":49,"nameZh":6,"description":50,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":51,"categories":52,"keywords":53,"publishedAt":19},"01M7875F3HQ1GF5F6BR36KY3JE","launchdeck-agent","Codex host adapter contract fixture for Launchdeck.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fagent-hosts\u002Fcodex\u002Fskill\u002FSKILL.md",[],[],{"skillNo":55,"name":49,"nameZh":6,"description":56,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":57,"categories":58,"keywords":59,"publishedAt":19},"01M7875DV2DQR33K1NFJHKAMZP","Claude host adapter contract fixture for Launchdeck.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002Ftest\u002Ffixtures\u002Fagent-hosts\u002Fclaude\u002Fskill\u002FSKILL.md",[],[],{"skillNo":61,"name":49,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M78759JQ7A6FQ8KDQTNYS3X9","Use this skill aggressively for local project lifecycle operations even when the user does not mention Launchdeck: start this project, run the dev server, dev, build, test, restart service, stop service, show logs, inspect ports, port is occupied, clean cache; and when the user explicitly asks to configure this local project for Launchdeck or create or generate a project-adapted `.launchdeck.yml`, including 为这个项目生成 Launchdeck 配置. Do not use for ordinary code edits, docs-only work, general explanations, API design discussion, or production deployment unless the user also asks to operate or configure a local project or service.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002Fagent\u002Finstaller-payload\u002Fskill\u002Flaunchdeck-agent\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M78753ZJQ7G2CZT1W1G8WRJV","zig-testing","Use ONLY when user explicitly requests Zig test code, built-in test runner configuration,\ntest failure diagnosis, or coverage for Zig projects.\n\nTriggers: \"write test zig\", \"zig test\", \"testing.allocator\", \"zig coverage\",\n\"comptime test\", \"zig expectError\", \"zig snapshot test\", \"zig fuzz\".\n\nDo NOT activate for: general Zig coding without test context, build.zig configuration\nfor non-test targets, Zig compiler errors, or Zig build system setup.","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fzig-testing\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":77,"licenseFileUrl":78,"sourceHtmlUrl":79,"categories":80,"keywords":81,"publishedAt":19},"01M78750JXY4Z5D5W1EX3PQ31Z","webapp-testing","Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.","Apache-2.0","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fwebapp-testing\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fwebapp-testing\u002FSKILL.md",[],[],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":19},"01M7874X2T83EQQFRCEY5SJJVC","verification-before-completion","Use when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always","https:\u002F\u002Fgithub.com\u002Fchenziyang110\u002Flaunchdeck\u002Fblob\u002Fmain\u002F.specify\u002Ftemplates\u002Fpassive-skills\u002Fverification-before-completion\u002FSKILL.md",[],[]]