Content
50%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-organized and highly actionable pytest reference, but it reads as a tutorial-style monolith rather than a lean skill: extensive sections cover basic pytest knowledge Claude already has, and no material is offloaded to reference files. The TDD workflow is sequenced but lacks explicit validation checkpoints. Trimming basics and splitting advanced topics into references would substantially improve token efficiency and structure.
Suggestions
Cut sections teaching pytest fundamentals Claude already knows (basic assertions, basic test structure, simple fixture/parametrize examples) and keep only project-specific conventions, retaining perhaps a quick-reference table.
Split advanced material — mocking catalog, async testing, and pytest configuration — into separate reference files (e.g. references/mocking.md, references/async.md) linked from a lean overview, enabling progressive disclosure.
Add explicit validation checkpoints to the TDD cycle, e.g. "RED: run pytest and confirm the new test fails before writing implementation; GREEN: run pytest again before refactoring."
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~800-line body spends large sections teaching pytest basics Claude already knows — an "Assertions" section covering equality, truthiness, membership, and comparisons, plus basic test structure (`assert 2 + 2 == 4`) and elementary fixture/parametrize usage — matching anchor 2 (noticeably verbose, several unnecessary explanations) rather than anchor 1 since the code blocks are terse and some sections (marker config, autospec, coverage config) add real value. | 2 / 5 |
Actionability | Provides concrete, mostly executable guidance throughout: copy-paste CLI commands (`pytest -m "not slow"`, `--cov` flags), complete pytest.ini/pyproject.toml configs, and concrete code patterns for fixtures, mocking, and parametrization. Some examples depend on undefined symbols (Database, create_app, mypackage) as templates, keeping it at anchor 4 rather than fully copy-paste-ready anchor 5. | 4 / 5 |
Workflow Clarity | The TDD cycle (RED/GREEN/REFACTOR) gives a clear numbered sequence, but validation checkpoints are implicit — it never instructs running pytest to confirm the failing test before implementing, or re-running after refactoring — matching anchor 3 (sequence present but checkpoints missing or implicit) rather than anchor 4. | 3 / 5 |
Progressive Disclosure | Sections are well-organized with clear headers, but this is a monolithic 800-line SKILL.md with no bundle files: async testing, the full mocking catalog, and configuration reference content that clearly belongs in separate reference files is fully inlined — matching anchor 3 (some structure, but content that should be separate is inline) rather than anchor 2 since navigation via headers is good. | 3 / 5 |
Total | 12 / 20 Passed |