← ClaudeAtlas

run-tddlisted

Execute the Leanforge Run workflow with selective TDD guidance. Use when the user invokes run-tdd, Leanforge:Run TDD, or asks to run Leanforge:Run while automatically applying TDD to observable behavior changes. This is a wrapper: Leanforge:Run remains the primary execution authority, and TDD is subordinate verification guidance for suitable implementation tasks.
pkkgh1519/leanforge · ★ 0 · Testing & QA · score 72
Install: claude install-skill pkkgh1519/leanforge
# Leanforge:Run TDD ## Purpose Use this skill as a thin wrapper around `Leanforge:Run` when the user wants Leanforge execution plus automatic TDD discipline. The original `Leanforge:Run` workflow remains authoritative. This skill only adds selective TDD guidance for implementation tasks that change observable behavior. This wrapper is self-contained and does not require a separately installed `tdd` skill. ## Required loading order 1. Read this `SKILL.md`. 2. Read the original `Leanforge:Run` `SKILL.md` completely and follow it as the primary workflow. 3. Apply the self-contained selective TDD policy below where this skill says TDD applies. If the original `Leanforge:Run` skill path is not obvious from available skill metadata, ask the user for the path instead of guessing. ## Precedence Follow this priority order: 1. System, developer, sandbox, tool, and safety instructions. 2. Project-local `CLAUDE.md`/`AGENTS.md` and repository constraints. 3. Original `Leanforge:Run` handoff/spec/plan execution workflow. 4. This wrapper's selective TDD guidance. If TDD guidance conflicts with the original `Leanforge:Run` flow, the Leanforge handoff/spec/plan execution flow wins. ## Selective TDD policy Apply TDD when a task changes observable behavior, such as: - business logic - public API behavior - state transitions - validation rules - error handling - parsing, calculation, sorting, filtering, or authorization behavior - user-visible flows For those tasks, use a vertical