tdd

name	tdd
description	Use for language-agnostic coder-agent behavior under strict test-driven development. Trigger when implementing a task with test-first workflow, RED/GREEN/REFACTOR cycles, minimal production changes, and verifiable ordering evidence for later evaluation.

TDD

Core Rule

Do not change production code until a failing test or executable check demonstrates the missing behavior.

Workflow

For each behavior slice:

1. RED: Add or select the smallest test/check for the next missing behavior. Run it and keep the failing output.
2. GREEN: Make the smallest production change that should pass the failing check. Run verification and keep the passing output.
3. REFACTOR: Refactor only after green. Run verification again.
4. Repeat until acceptance criteria are met.

Constraints

• Use the stack and commands from the provided prompt file or task context.
• Keep cycles small; avoid broad rewrites.
• Test observable behavior, not implementation details.
• Preserve ordered evidence for the evaluator.
• Do not claim TDD compliance unless the artifacts show it.

Final Response

Report only:

• what behavior was completed
• verification commands and results
• where the evidence is located
• any known gaps

When running inside this benchmark, package the final evidence using $eval-result.