AgentSkillsCN

code-quality-testing

践行测试效率与有效性原则。适用于编写测试用例、评审测试套件，或提升测试覆盖率时使用。重点聚焦行为覆盖与变异感知型断言。

SKILL.md
--- frontmatter
name: code-quality-testing
description: Apply testing efficiency and effectiveness principles. Use when writing tests, reviewing test suites, or improving test coverage. Emphasizes behavioral coverage and mutation-aware assertions.

Code Quality: Testing

Tests are production code. Apply the same quality standards.

Structure

  • •Every test tests one behavior. A failing test should tell you exactly what broke without reading the test body.
  • •Test names are sentences: test_chunker_splits_on_sentence_boundary_not_word_boundary is better than test_chunker_2.
  • •Test docstrings: "As a user I want [objective], so I can [benefit]." for user-facing tests. For technical/utility tests, use full docstring with Args/Returns or a concise one-liner. See test-documentation.mdc.

Efficiency

  • •Prefer narrow, fast unit tests. Mock external deps (Ollama, ChromaDB, auth) at the boundary — not deep inside logic.
  • •Mark slow/external tests. Any test touching network, Docker, or Ollama gets the right marker (@pytest.mark.slow, @pytest.mark.ollama, etc.).
  • •Avoid duplication. If several tests set up the same fixture with minor variations, use @pytest.mark.parametrize.

Effectiveness

  • •Mutation testing (mutmut) guidance: Assert specific outcomes. An assertion that passes when the condition is inverted doesn't test anything. Each test must assert a definite result, not just "no exception raised".
  • •Coverage is a floor, not a goal. 100% coverage with weak assertions is worse than 80% with assertions that catch mutations. Prioritize behavioral coverage over line coverage.