TDD
Test-Driven Development (TDD) is a software development methodology whose core cycle consists of writing a failing unit test first, then writing the minimal implementation code to make the test pass, and finally refactoring. The method uses short iterative loops as its basic unit, emphasizes writing tests before code, and aims to improve code quality, maintainability, and design soundness. It is widely applied in agile software development practices.
Core Facts
Test-Driven Development (TDD) was proposed by Kent Beck in 1999, with a core workflow of red-green-refactor.
The core TDD cycle is: first write a failing test case (Red) → write the minimum amount of code to make the test pass (Green) → refactor the code (Refactor).
SDD（Specification-Driven Development，规范驱动开发）是当前AI编程领域被普遍认可的一种开发思想，核心理念是基于规范驱动进行编程
Timeline (last 90 days)
TDD moves acceptance criteria forward, which can reduce cases of "looks fine but actually has defects" and lets regression issues surface early.
The implement spec skill packages together the TDD skill, the code review skill, and multi-worktree parallelism: it first organizes tickets and identifies tickets with no dependencies, then launches multiple AI agents simultaneously to work in parallel.
The difference between PStack's Create Verification Skill in the testing phase and TDD is that TDD focuses on functional tests and unit tests, whereas this skill focuses on building an end-to-end verification environment.
Pstack's build phase includes a workflow that follows TDD to write code, as well as a best-practice skill that forces the AI to write code according to TypeScript conventions.
Matt Pocock strongly recommends test-driven development (TDD) for AI agents; writing the failing test first makes it hard for AI to generate weak tests that merely confirm its own existing implementation.
The executing party uses the Superpowers plugin (version 6.3.0), both AIs have the same version installed, and the most frequently requested skills include writing requirement specifications, executing according to plan, red before green (TDD), and verification before completion.
The core of Pstack's build phase is test-driven development (TDD), and it includes built-in TypeScript best-practice skills.
The method behind the proof gate is test-first, originating from TDD's red-green-refactor cycle, and is used to verify that the failure of a new test proves it actually covers the behavior to be fixed rather than being a test that always passes vacuously.
Implement Spec dispatches a SubAgent responsible for implementation to handle each Ticket in a TDD manner.
The workflow combining TDD with AI Agent is: first write test cases, have the Agent generate implementation code that passes the tests, and after the Agent executes the tests it automatically iterates and revises based on failure information until all tests pass.
7 more timeline events
All Facts (20)
Test-Driven Development (TDD) was proposed by Kent Beck in 1999, with a core workflow of red-green-refactor.
80%VerifiedThe core TDD cycle is: first write a failing test case (Red) → write the minimum amount of code to make the test pass (Green) → refactor the code (Refactor).
80%VerifiedSDD（Specification-Driven Development，规范驱动开发）是当前AI编程领域被普遍认可的一种开发思想，核心理念是基于规范驱动进行编程
80%VerifiedTest-Driven Development (TDD) was systematically proposed by Kent Beck within the Extreme Programming (XP) methodology.
80%VerifiedA TDD test-driven approach can be adopted, having the AI write tests first and then write the implementation to pass those tests.
65%UnverifiedSuperpowers的TDD流程要求AI先编写失败的测试用例，再编写刚好能让测试通过的最小实现代码
95%UnverifiedSuperpowers引入了TDD（测试驱动开发）和YAGNI（You Aren't Gonna Need It）两大软件工程原则
95%UnverifiedTDD（Test-Driven Development）由Kent Beck在2003年正式提出
90%Unverified验证合约的设计借鉴了测试驱动开发（TDD）和契约式设计（Design by Contract）两大工程传统
90%UnverifiedHarness Engineering通过强约束沙盒和严苛TDD模式来解决智能体漂移问题
80%Unverified测试驱动开发（TDD）在AI编程时代变得更加重要，因为它提供了不依赖对AI信任的客观验证机制
75%UnverifiedThe rule in CLAUDE.md to 'first write a test that reproduces the problem, then make it pass' is essentially a migration of Test-Driven Development (TDD) philosophy into the AI collaboration context.
60%UnverifiedTDD moves acceptance criteria forward, which can reduce cases of "looks fine but actually has defects" and lets regression issues surface early.
50%UnverifiedThe implement spec skill packages together the TDD skill, the code review skill, and multi-worktree parallelism: it first organizes tickets and identifies tickets with no dependencies, then launches multiple AI agents simultaneously to work in parallel.
50%UnverifiedThe difference between PStack's Create Verification Skill in the testing phase and TDD is that TDD focuses on functional tests and unit tests, whereas this skill focuses on building an end-to-end verification environment.
50%UnverifiedPstack's build phase includes a workflow that follows TDD to write code, as well as a best-practice skill that forces the AI to write code according to TypeScript conventions.
50%UnverifiedMatt Pocock strongly recommends test-driven development (TDD) for AI agents; writing the failing test first makes it hard for AI to generate weak tests that merely confirm its own existing implementation.
50%UnverifiedThe executing party uses the Superpowers plugin (version 6.3.0), both AIs have the same version installed, and the most frequently requested skills include writing requirement specifications, executing according to plan, red before green (TDD), and verification before completion.
50%UnverifiedThe core of Pstack's build phase is test-driven development (TDD), and it includes built-in TypeScript best-practice skills.
50%UnverifiedThe method behind the proof gate is test-first, originating from TDD's red-green-refactor cycle, and is used to verify that the failure of a new test proves it actually covers the behavior to be fixed rather than being a test that always passes vacuously.
50%