TDD Loop Master

TDD Loop Master enforces the Red-Green-Refactor cycle as a strict gate system, not a suggestion. It opens every session by turning your requirement into a ranked test list — degenerate null cases first, complex edge cases last — before any code exists. From there it controls the pace: one failing test, the minimum production code to make it pass, then a refactor window only once the suite is green. It also audits any mocks you bring in, checking whether they could conceal a real failure.

A typical session starts with something like: 'I need to build a discount-calculation function in Python using PyTest. The rules are tiered by order value and customer status.' The skill asks four short context questions — stack, goal, constraints, audience — then generates the opening test list and writes the first forcing test in PyTest syntax, explaining exactly why it must fail before you touch production code.

The output for that first cycle might look like: Test List (excerpt): – [ ] calculate_discount(0, 'standard') returns 0.0 – [ ] calculate_discount(100, 'standard') returns 0.0 – [ ] calculate_discount(100, 'vip') returns 10.0 – [ ] calculate_discount(-1, 'standard') raises ValueError Red Phase — Forcing Test: def test_zero_order_returns_no_discount(): assert calculate_discount(0, 'standard') == 0.0 # Function does not exist yet — this must fail to proceed. Green Phase instruction: implement the narrowest code that passes only this assertion.