Skip to content

DevNullLtd/Yuzu / .codex/skills/tdd

Apply Yuzu-specific test-driven development using tracer-bullet red-green-refactor cycles and the repo's Meson, gateway, proto, and ABI validation rules. Use when the user says `/tdd`, asks for TDD, red-green-refactor, or wants a feature/fix built test-first in Yuzu.

TDD

Use behavior-first tests through public interfaces. Prefer integration-style seams that exercise real stores, REST handlers, gateway paths, plugin boundaries, or command dispatch contracts. Avoid tests coupled to private implementation.

Before Red

  • Read CONTEXT.md and relevant ADRs in docs/adr/.
  • Identify the public interface and the project term for the behavior.
  • List the smallest vertical behavior slice that proves value.
  • Choose the narrowest validation path through $yuzu-build.
  • Invoke $yuzu-proto, $yuzu-plugin-abi, $yuzu-meson, or $yuzu-windows-msvc if the slice crosses those boundaries.

Cycle

  1. Red: add one failing test for one observable behavior.
  2. Green: implement only enough code to pass that test.
  3. Verify: run the targeted suite with meson test -C <builddir> --suite <suite> --print-errorlogs or the relevant gateway command.
  4. Repeat for the next behavior.
  5. Refactor only while green.

Do not write all tests first. Use tracer bullets: one behavior, one test, one implementation.

Yuzu Test Hygiene

  • Use yuzu::test::unique_temp_path(prefix) or TempDbFile for C++ test temp files and SQLite DBs.
  • Avoid hardcoded shared paths and clock/thread-hash uniqueness.
  • For gateway EUnit, pass --dir apps/yuzu_gw/test.
  • For direct Erlang isolation, remove the relevant _build/test tree or run in a fresh BEAM.
  • For proto changes, include buf lint proto and buf breaking proto --against '.git#subdir=proto,ref=origin/main' when available.
  • For plugin ABI changes, add descriptor/loader coverage and preserve stable C ABI compatibility.

Done

Finish with the tests added, implementation green, targeted validation output summarized, and any broader /test --quick or governance follow-up called out.

finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 319kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 319kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 319kUpdated Claude CodeCodexVerifiedInstall
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