AGENT SKILL

principles

Answers test design questions from authoritative TDD sources (Beck, Khorikov), producing WHY reasoning to improve test design decisions. Use when: 'should I mock this', 'four pillars of a good test', 'red green refactor', 'classical vs london school', 'test doubles', 'testing anti-patterns', 'resistance to refactoring', 'code coverage', 'observable behavior', 'humble object', 'integration test', 'test pyramid', 'output vs state vs communication test', 'property-based testing', 'characterization test', 'legacy code'. Not for HOW to run tests in the project (use the project's own test tooling and workflow for that).

License 许可
MIT
最近更新
2026-11-15
Stars
★ 30
Forks
7

01 前言 Front Matter

metadata
{ "summary": "Answer test design questions from authoritative TDD sources", "workflow-stage": "test" }
description
Answers test design questions from authoritative TDD sources (Beck, Khorikov), producing WHY reasoning to improve test design decisions. Use when: 'should I mock this', 'four pillars of a good test', 'red green refactor', 'classical vs london school', 'test doubles', 'testing anti-patterns', 'resistance to refactoring', 'code coverage', 'observable behavior', 'humble object', 'integration test', 'test pyramid', 'output vs state vs communication test', 'property-based testing', 'characterization test', 'legacy code'. Not for HOW to run tests in the project (use the project's own test tooling and workflow for that).
argument-hint
[question or concept]
user-invocable
true
disable-model-invocation
false

02 资源文件 16

03 SKILL.md 正文

TDD Knowledge Base

The Beck and Khorikov files are distilled from cover-to-cover reading of those books. Two files are not: property-based-testing.md and characterization-testing.md are built from web sources, listed separately under Sources. Reference files in reference/ have author-attributed sections with synthesis where authors overlap.

Routing Table

Query about... Load
TDD cycle, red-green-refactor, when/why TDD, step size, courage, stress, fear methodology-beck.md
What makes a good test, what to test, 3A, Fixture, test naming, parameterized tests test-design.md
Mocks, stubs, fakes, mock/stub taxonomy, CQS connection, 5 mocking best practices test-doubles.md
Refactoring safely, design patterns in TDD, Humble Object, four types of code refactoring-under-test.md
Four Pillars of a good test, brittle tests, false positives, resistance to refactoring four-pillars-khorikov.md
Code coverage metrics, branch vs statement, goal of unit testing code-coverage-khorikov.md
Classical vs London school, dependency taxonomy, shared/private/out-of-process classical-vs-london-khorikov.md
Observable behavior, implementation details, hexagonal architecture, encapsulation observable-behavior-khorikov.md
Output/state/communication styles, functional architecture, functional core testing-styles-khorikov.md
Four types of code 2x2, Humble Object, CRM refactoring, CanExecute/Execute, domain events, preconditions testable-architecture-khorikov.md
Integration tests, managed vs unmanaged deps, Test Pyramid, database testing, logging, interfaces, reads vs writes integration-testing-khorikov.md
Anti-patterns: private methods, exposing state, domain knowledge, code pollution, DateTime.Now, time anti-patterns-khorikov.md
Beck's Money example, Value Object, Expression metaphor, Factory Method money-example-beck.md
Beck's xUnit example, bootstrap, Template Method, Composite, Collecting Parameter xunit-example-beck.md
Property-based testing: properties as specification, shrinking, property patterns, properties beside example tests, agent-written properties property-based-testing.md (web-sourced)
Legacy code, characterization tests, golden master, approval and snapshot tests, sabotage check, pins prove sameness not correctness characterization-testing.md (web-sourced)

Load the most relevant file first. Load a second only if the first doesn't fully answer.

Quick decision guide (no file load needed):

  • "Should I write a test for this?" → Yes, unless it's third-party code you trust or trivial code (bottom-left quadrant of the 2x2 matrix: low complexity, few collaborators)
  • "Should I mock this?" → Only if it's an unmanaged dependency (message bus, SMTP). Use real instances for managed deps (database)
  • "Fake It or Obvious Implementation?" → If confident, Obvious. If surprised by red, back off to Fake It
  • "Is this test too big?" → If it needs >3 changes to work, write a smaller Child Test first
  • "Test before or after?" → Before. Always. "You won't test after"
  • "Output, state, or communication test?" → Output-based first. State-based second. Communication-based only for inter-system with visible side effects
  • "Is this code testable?" → Check the 2x2 matrix. High complexity + many collaborators = split it (Humble Object)
  • "Should I introduce an interface?" → Only for unmanaged deps you need to mock. Concrete classes for managed deps. Never for in-process deps
  • "Should I test this precondition?" → Yes if it has domain significance (e.g., non-negative employees). No if it's purely technical (array length check)
  • "Should I test this read operation?" → Higher threshold than writes. Writes corrupt data, so always test. Reads: only the most complex/important ones
  • "Domain event or direct call?" → Use domain events when the domain model needs to trigger external notifications without depending on out-of-process deps
  • "Why is my test brittle?" → It's probably coupled to implementation details. Check: are you asserting on observable behavior or internal structure?

Sources

  • Beck: Kent Beck, Test-Driven Development: By Example (2003)
  • Khorikov: Vladimir Khorikov, Unit Testing: Principles, Practices, and Patterns (2020)
  • Ousterhout (secondary, cross-referenced only): John Ousterhout, A Philosophy of Software Design, cited in one editorial-synthesis section of test-doubles.md
  • Mutation-testing literature (secondary, cross-referenced only): the primary sources on mutation testing, cited in one clearly-labeled editorial note in each of code-coverage-khorikov.md and four-pillars-khorikov.md, where a claim of Khorikov's has a measurable partial exception he does not cover. Both notes are marked as outside his text and leave his claim standing. The material itself is owned by /mutation-testing:principles when the mutation-testing plugin is enabled; without it, each note carries its own one-line fallback.

Web-sourced references (not book readings)

These two files were built from web pages, papers' abstract and HTML pages, and tool docs read on 2026-10-06, not from reading a book. Each file lists its sources with read depth and a recheck trigger; this list is rechecked when either file's trigger fires.

  • property-based-testing.md: Hypothesis, fast-check and other property tools' docs; Maaz et al. (arXiv 2510.09907, NeurIPS 2025 workshop paper); Vikram et al. (arXiv 2307.04346); PROBE (ACL 2026 Findings). No property-based testing book was read.
  • characterization-testing.md: Michael Feathers' 2016 "Characterization Testing" page (his book Working Effectively with Legacy Code was not read); Mark Seemann's 2025 "Empirical characterization testing" post; ApprovalTests, Verify and Jest docs.

Naming convention

  • {concept}.md is shared across authors, with attributed sections inside
  • {concept}-{author}.md is used when only one author covers it substantively

Next

/testing:write

Puts the test-design reasoning into the tests themselves.

Scope boundary

This skill is knowledge (WHY behind testing decisions), not workflow (HOW to run tests). For running, filtering, or scaffolding tests, use your project's own test tooling, conventions, and workflow skills. This skill informs the design decisions those workflows execute.

04 同仓库更多 Skills

melodic-software
melodic-software/claude-code-plugins

track

Backlog reads through the bound tracker: search open and closed items, view one item with its comments and state history. Use when another skill searches or reads tracker items.

★ 30
melodic-software
melodic-software/claude-code-plugins

reviewer-media

Reviewer media for one pull request: decides whether it needs a narrated walkthrough, builds the storyboard from /review:explain-change's digest, has /explainer-video:produce render it and hands the manifest to /source-control:pull-request media. Use when asked for 'reviewer media', 'a walkthrough video for this PR', 'a produced video of this change', or when pull-request ready or monitor hands a PR over.

★ 30
melodic-software
melodic-software/claude-code-plugins

route

Model and effort per agent role (orchestrator, worker, verifier, retrieval) for a multi-agent run, resolved through the bundled defaults and the user, team and personal override config layers into fenced JSON with single-agent and fan-out variants and the layer behind each value. Use when asked 'which model for the subagents', 'route the roles' or 'model routing', or before a skill passes a role's model on a spawn.

★ 30
melodic-software
melodic-software/claude-code-plugins

route

Model and effort per agent role (orchestrator, worker, verifier, retrieval) for a multi-agent run, resolved through the bundled defaults and the user, team and personal override config layers into fenced JSON with single-agent and fan-out variants and the layer behind each value. Use when asked 'which model for the subagents', 'route the roles' or 'model routing', or before a skill passes a role's model on a spawn.

★ 30
melodic-software
melodic-software/claude-code-plugins

principles

Module-design vocabulary for architecture findings, and the checks built on it. Use when: 'what is a deep module', 'what is a seam', 'deep vs shallow module', 'what is the deletion test', 'leverage and locality', 'module design vocabulary', 'which word do we use for X in architecture findings'. Scanning a codebase for friction is /architecture:improve.

★ 30
melodic-software
melodic-software/claude-code-plugins

write

Structured five-field bug report (title, steps to reproduce, expected vs actual, justified severity, suggested fix location) from an informal description, read-only. Use when the user names a defect they observed, or asks for one written up ('report a bug', 'file a bug', 'write this up as a bug').

★ 30