Skip to content

TsukiNi22/ai-utils / skills/tests/tests

Set up and write unit tests for a project in any language - detects the languages and build system, reuses the test framework already in place or asks the user which framework/method to use (GTest, Catch2, doctest, Criterion, pytest, vitest/jest, cargo test, go test, bats...), then sets it up, inventories what to test, writes test files mirroring the source tree, runs them and adds the CI step. For C++ it follows cpp-tests. Use whenever the user wants unit tests, a test setup, more coverage or "test the whole project" and the project is not (only) C++.

Unit tests (any language)

C++ (CMake) project or C++ part of a project: load and follow cpp-tests (GoogleTest, libutils way). Everything else follows this file.

1. Detect

  • Languages / build: CMakeLists.txt, Makefile, *.c, package.json (+ lock file), pyproject.toml / setup.cfg / requirements*.txt, Cargo.toml, go.mod, *.sh, *.lua...
  • Existing tests: tests/, test/, __tests__/, *_test.go, test_*.py, *.spec.ts, framework in the dependencies, ctest/pytest/vitest config, CI step running tests.
  • A framework is already used: reuse it, say so, don't ask.

2. Ask (AskUserQuestion, French) when nothing is in place

One question per language to set up, the default first with "(Recommandé)":

LanguageOptions (default first)
C++GoogleTest (Recommandé, user's choice in libutils) · Catch2 · doctest
CCriterion (Recommandé) · CUnit · Unity
Pythonpytest (Recommandé) · unittest
JS / TSvitest (Recommandé for Vite/ESM) · jest · node:test
Rustcargo test (built-in) · + proptest
Gogo test (built-in) · + testify
Shellbats-core (Recommandé) · shunit2

Also ask, in the same call when relevant: scope (whole project / given modules), and whether the CI must run them. Don't install anything system-wide without confirmation (package managers: give the command; root -> graphical sudo -A, see the global CLAUDE.md).

3. Set up

  • Add the dependency the way the project manages them (dev dependency, CMake find_package, Cargo.toml [dev-dependencies]...), the config file only if needed, a test entry point (npm test, make tests, ctest, pytest), the produced files in .gitignore (coverage, caches, test binaries).
  • CI: one step/job that builds and runs the tests on push (in the project's existing workflow when there is one; git-conventions / cpp-project workflows for the user's projects).

4. Write

  • Inventory first: list the public modules / classes / functions and which ones have tests; for a big project show the plan (files + main cases) before writing.
  • One test file per source file/module, mirroring the source tree (src/net/client.py -> tests/net/test_client.py, src/parser.ts -> src/parser.test.ts or tests/parser.test.ts per the framework convention).
  • Per public unit: default state, nominal behavior, edge cases, errors (exact error type/code/message), concurrency / timing with timeouts (never fixed sleeps), anything that can hang isolated (sub-process / timeout option of the framework), global state restored after each test, fakes/mocks only at the boundaries (network, files, time, processes).
  • Test names say the behavior (test_cancel_unknown_id_raises, it("returns null on empty input")).
  • Tests describe the expected behavior: a test revealing a bug stays failing and is reported; never change the production code to make a test pass unless the user asks.
  • Comments and style of the test code: the user's comment style (comments / cpp-comments) and the conventions of the project.

5. Run & report

Run the whole suite (and the new tests alone), report: tests added per module, passed / failing (= bugs found, with the reason), what remains untested, how to run them (<command>), CI added or not. Deliver the report through the report skill (ask the output format of report (Markdown + PDF recommended, PDF only, Markdown only)), in English, in audit/ at the root of the repository (or of the current folder outside a repository), never docs/ unless asked (audit/<YYYY-MM-DD>-tests.*), not only in the chat.

finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 318kUpdated Claude CodeCodexUnverifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 318kUpdated Claude CodeCodexUnverifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 318kUpdated Claude CodeCodexUnverifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 196kUpdated Claude CodeCodexVerifiedInstall