跳到正文
cccskills
免费在 GitHub 公开

mull-mutation

Runs Mull, the LLVM-IR mutation testing tool, against C/C++ test binaries built with Clang: covers install (the version-matched mull-NN package), the -fpass-plugin build flags for the Mull IR frontend, mull-runner invocation, the mutator catalog, path filtering, and GitHub Actions CI. Use when a C or C++ project needs mutation-score verification with the tool already chosen. Does not select among mutation tools and does not cover other languages (stryker-mutation for JS/TS, stryker-net-mutation for .NET, pitest-mutation for the JVM, mutmut-mutation for Python).

中文简介正在准备中。当前显示原文说明。

testland/qa星标 13(GitHub 13 + cccskills 0)
集成工具GitHub

实测评分 未实测

查看安装方法

包含的文件(2)

  • SKILL.md8.7 KB
  • .tessl-plugin/plugin.json697 B

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

mull-mutation

Overview

Per mull-readme:

"Mull is a practical mutation testing and fault injection tool for C and C++."

Per mull-readme, Mull is "LLVM-based" and "Built on LLVM IR (Intermediate Representation) and utilizes LLVM JIT compilation."

The LLVM-IR foundation means Mull operates at a layer below the source language - any LLVM-emitting frontend (Clang, Rust's rustc, Swift's swiftc) can in principle work, though C / C++ are the primary targets.

The foundational paper per mull-readme: "Mull It Over: Mutation Testing Based on LLVM" (2018), Denisov & Pankevich.

When to use

  • A C / C++ project needs mutation-quality verification of its test suite.
  • A library shipping safety-critical code (embedded, kernel, cryptography) needs the highest-confidence test suite.
  • Language ecosystems without a native mutation tool need an LLVM-IR-level fallback.

Step 1 - Install

Latest release per mull-changelog: 0.34.0 (May 2026). Mull ships as a version-matched package: mull-19 bundles LLVM 19.1.7, and the suffix MUST match the Clang you compile with (plugin / ABI compatibility), per mull-install.

# macOS / Linuxbrew
brew install mull-project/mull/mull

# Linux (Ubuntu/Debian): install the version-matched mull + clang packages
# from the Mull apt repo per the official guide: https://mull.readthedocs.io

Step 2 - Build the project for mutation

Mull needs the test binary built with -O0 -g (no optimization, debug symbols) and the Mull IR frontend, which current Mull loads as a compiler plugin via -fpass-plugin (the old -Xclang -load legacy-pass approach is gone), per mull-tutorial. The plugin path is version-suffixed and must match the installed mull-NN package and the Clang version:

# CMake-based project. The "19" in the plugin path, the mull-19 package, and
# clang-19 must all agree.
cmake -DCMAKE_C_COMPILER=clang-19 -DCMAKE_CXX_COMPILER=clang-19 \
      -DCMAKE_C_FLAGS="-O0 -g -grecord-command-line -fpass-plugin=/usr/lib/mull-ir-frontend-19" \
      -DCMAKE_CXX_FLAGS="-O0 -g -grecord-command-line -fpass-plugin=/usr/lib/mull-ir-frontend-19" \
      -B build/

cmake --build build/

The Mull pass instruments the compiled IR with conditional mutation points; at runtime, the runner enables / disables each mutation to test which the test suite catches.

Step 3 - Run

# The runner binary is version-suffixed too. The default IDE reporter prints
# killed / survived mutants to the console (per [mull-cli]).
mull-runner-19 build/tests/MyTests

Output:

[killed]   src/cart.cpp:42 - Conditional Boundary
[survived] src/cart.cpp:78 - Arithmetic Operator Replacement
[survived] src/cart.cpp:103 - Statement Removal

Mutants killed: 87 (74.4%)
Mutants survived: 30
Mutation score: 74.4%

Step 4 - Mutators

Mull's mutator catalog (representative; per the paper + code):

MutatorExample
Arithmetic Operator Replacement+ → -
Conditional Boundary< → <=
Conditional Negation== → !=
Statement Removaldelete x; → (removed)
Return Value Replacementreturn x; → return 0;
Constant Mutation42 → 0, 1 → 0

LLVM-IR-level mutation means some source-level patterns produce equivalent IR (no mutation possible) - Mull skips those silently.

Step 5 - Filtering

Mull accepts include / exclude path filters:

mull-runner-19 build/tests/MyTests \
  --include-path "src/checkout/*" \
  --exclude-path "third_party/*" \
  --exclude-path "tests/*"

Always exclude tests + third-party - mutating those is meaningless.

Step 6 - Reports

# Elements = interactive HTML report (mutation-testing-elements);
# Sarif = machine-readable for GitHub Code Scanning. Per [mull-cli], pick
# reporters by name and set the output dir / base name explicitly.
mull-runner-19 build/tests/MyTests \
  --reporters Elements,Sarif --report-dir ./mull-report --report-name results

# Outputs under ./mull-report/ (results.html, results.sarif)

The Elements report shows a per-file mutation breakdown; the Sarif output uploads to a code-scanning dashboard. Other reporters per mull-cli: IDE (console), SQLite, GithubAnnotations, Patches.

Step 7 - CI integration

- run: |
    # install mull-19 + clang-19 per the official guide (Step 1) before this step
    cmake -B build/ -DCMAKE_BUILD_TYPE=Debug \
      -DCMAKE_C_COMPILER=clang-19 -DCMAKE_CXX_COMPILER=clang-19 \
      -DCMAKE_C_FLAGS="-O0 -g -fpass-plugin=/usr/lib/mull-ir-frontend-19" \
      -DCMAKE_CXX_FLAGS="-O0 -g -fpass-plugin=/usr/lib/mull-ir-frontend-19"
    cmake --build build/
    mull-runner-19 build/tests/MyTests --reporters Sarif --report-dir ./mull-report --report-name results
- uses: actions/upload-artifact@v4
  if: always()
  with:
    name: mull-results
    path: ./mull-report/

For PR-incremental runs, scope to changed files:

CHANGED=$(git diff --name-only origin/main...HEAD | grep -E '\.(c|cpp|h|hpp)$')
mull-runner-19 build/tests/MyTests --include-path "$(echo $CHANGED | tr ' ' ',')"

Anti-patterns

Anti-patternWhy it failsFix
Mutating Release buildsOptimizer changes IR shape; mutators behave unpredictably.Build with -O0 -g (Step 2).
Including tests/ in mutationMutates the tests; meaningless.Exclude tests path (Step 5).
Mismatched LLVM versions (mull-19 plugin but compiled with clang-17)Plugin load / ABI incompatibility.Match the suffix across package, plugin path, and clang (Step 2).
Running on every PRSlow; team disables.Schedule + per-changed-file scope (Step 7).
Setting an unrealistic mutation score gateForces low-value tests.Start at current baseline.

Limitations

  • LLVM toolchain dependency. Requires the same LLVM version Mull was built against; can be tricky in heterogeneous CI.
  • C / C++ primary; other languages incidental. Swift / Rust via LLVM-IR work in principle but with less polish.
  • Build configuration constraint. -O0 builds run slower than Release; mutation-runs are inherently slow.
  • No native incremental mode. Use --include-path for git-diff scoping (Step 7).
  • Smaller community than Stryker / PIT. 811 stars per mull-readme vs Stryker's thousands; tooling integration varies.

References

  • mr - Mull README: LLVM-based, C/C++ primary, foundational 2018 paper "Mull It Over: Mutation Testing Based on LLVM" (Denisov & Pankevich).
  • mull-changelog - release history (latest 0.34.0, May 2026).
  • mull-install - apt repo + version-matched mull-NN package.
  • mull-tutorial - -fpass-plugin build instrumentation.
  • mull-cli - mull-runner flags and reporter names.
  • stryker-mutation, stryker-net-mutation, pitest-mutation, mutmut-mutation - per-language siblings.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Builds a CI gate that fails the build on **new** WCAG / a11y violations introduced by a PR while grandfathering pre-existing violations on a per-rule / per-page baseline. Aggregates verdicts from axe-core / pa11y / Lighthouse a11y / WAVE / IBM Equal Access scans. Use when a project has accumulated a11y debt and a strict "zero violations" gate would block every PR - the ratchet pattern lets the team ship while preventing regressions.

中文简介正在准备中。当前显示原文说明。

testland/qa132026年10月20日 更新

Workflow skill that builds an A/B-test validity checklist from an experiment proposal, walking the canonical design-correctness gates - pre-registered OEC/power/guardrails, randomization unit + SRM check, assignment integrity, telemetry, peeking discipline, novelty/primacy, post-experiment SRM re-check - into a per-experiment checklist + sign-off form. Use when launching, auditing, or governing an experiment. For pitfall mechanics (guardrails, peeking) see experiment-results-interpreter's references; to read an already-valid result use experiment-results-interpreter; for per-SDK harness tests use experiment-sdk-testing - this gates DESIGN, not SDK code.

中文简介正在准备中。当前显示原文说明。

testland/qa132026年10月20日 更新

The single install-flow skill: the reference contract for the PWA install flow (installability gate fields, `beforeinstallprompt` handshake, per-platform paths - Android WebAPK / iOS Share menu / Firefox no-op - and the `display-mode` post-install signal, in references/install-flow-reference.md) plus the build-an-X workflow that emits the Add-to-Home-Screen suite. Walks the four-stage timeline, emitting one test per gate cell per [web.dev/articles/install-criteria][install-criteria], the deferred-prompt → `prompt()` → `userChoice` chain per [web.dev/articles/customize-install][customize-install], the iOS Safari manual-metadata branch (`apple-touch-icon`) per [web.dev/learn/pwa/installation][learn-pwa], and the post-install `(display-mode: standalone)` MQ assertion. Output: a Playwright spec with per-stage cells plus a coverage matrix. Use when a PWA's manifest, icons, or install handler change, when install conversion drops at an unknown stage, or when triaging a flaky install assertion.

中文简介正在准备中。当前显示原文说明。

testland/qa132026年10月20日 更新

Build-an-X workflow that uses an LLM to map existing tests to spec sections - given a spec doc + the test suite, the LLM identifies which tests cover which sections, surfaces uncovered sections (gap), and recommends specific tests to add. Output is a coverage matrix per spec ID. Scope is mapping tests that already exist and naming the gaps, not authoring tests for new acceptance criteria. Use when a spec doc and a test suite both exist but nobody can say which requirements are actually covered - before a release sign-off, an audit, or a decision about where to spend the next round of test effort.

中文简介正在准备中。当前显示原文说明。

testland/qa132026年10月20日 更新

Generates tests from natural-language specs (acceptance criteria, user stories, requirements) using an LLM, with confidence scoring per test case (LLM self-assessment plus heuristics: assertion quality, naming, completeness), batching uncertain cases for human review, and integration with the team's existing test framework. Use when the user asks to generate unit tests from acceptance criteria, convert user stories to test cases, automate test creation from requirements, or augment a spec-driven test suite with AI-generated stubs that are then curated before merge.

中文简介正在准备中。当前显示原文说明。

testland/qa132026年10月20日 更新

Configures Allure Report (test-runner adapter install, `allure-results` directory wiring, `categories.json` for failure classification, `history-trend.json` retention via the copy-history-between-runs pattern), runs the Allure CLI to convert `allure-results` to a static HTML site, and uploads the report as a CI artifact. Use when the team needs richer test reporting than JUnit XML - step-level attachments, per-test history, retry tracking, and severity / epic / feature labeling across framework-agnostic adapters (pytest, Jest, JUnit, TestNG, NUnit, Mocha). As a rich static HTML report generator, it is the open-source alternative to the sunset ExtentReports (JVM/.NET per-test HTML narrative); for hosted cross-run flakiness analytics rather than a static per-run report use currents-integration.

中文简介正在准备中。当前显示原文说明。

testland/qa132026年10月20日 更新

查看 testland 的全部技能

报告此技能的问题