tdd-execution

Test-driven development execution automates the RED-GREEN cycle by spawning tester agents to write tests first, then coder agents to implement code that passes those tests. This Skill coordinates that workflow across user stories with acceptance criteria, enabling parallel processing of up to 10 concurrent tester-coder pairs while maintaining sequential RED→GREEN phases within each story.

Yes. This Skill organizes work into waves where each wave contains multiple user stories. Within each story, RED and GREEN phases run sequentially, but across different stories in the same wave, tester-coder pairs execute in parallel—up to 10 concurrent pairs—for faster delivery.

User stories must include acceptance criteria and an execution plan structured as waves. The Skill then spawns agents to generate tests and implementations, tracks status through metadata updates, and verifies all tests pass before marking completion.

Agent orchestration automates role separation: dedicated tester agents write comprehensive tests during the RED phase, while dedicated coder agents implement features to pass those tests during GREEN. This reduces context switching and ensures tests genuinely drive design before code is written.

The Skill updates user stories with test and implementation metadata for audit trails, verifies all tests pass against the generated code, and marks the story complete only after both phases succeed—providing visibility into what was tested and implemented.

Yes. Wave-based orchestration scales across multiple concurrent tester-coder pairs (up to 10), making it suitable for delivering multiple user stories in parallel while maintaining test-first discipline and sequential RED→GREEN integrity per story.