Tdd

Implement $ARGUMENTS using strict TDD.

Iron Laws

> 1. No production code exists without a failing test. > If you write code before a test, DELETE it. Not "keep as reference." Not "adapt it." DELETE it. > > 2. Write the MINIMUM code to pass the test. > If the test passes, STOP writing code. More code = more bugs = more maintenance. > > 3. Delete means delete. > If you wrote code before the test: don't keep it as reference, don't paste it somewhere, don't look at it. Delete it and write the test first. Your memory of the solution will make reimplementation fast.

Cycle

1. RED — Write a Failing Test

• Find existing test files to match the naming pattern
• Write ONE test that describes ONE specific behavior
• Run it — confirm it FAILS

`` [test command] ``

• If it passes without new code, your test is wrong — it's testing something that already exists

2. GREEN — Write Minimum Code

• Write the simplest possible code that makes the test pass
• No extra features, no "while I'm here" additions, no premature optimization
• Run the test — confirm it PASSES

`` [test command] ``

3. REFACTOR — Clean Up

• Improve code quality without changing behavior
• Remove duplication, improve naming, simplify logic
• Run ALL tests — confirm nothing broke

`` [test command] ``

4. REPEAT

If the feature needs more behavior:

• Go back to step 1 with the next test case
• Continue until all requirements from $ARGUMENTS are covered
• Each cycle should be small: one test, one behavior

Anti-Rationalization

| Excuse | Rebuttal | |--------|----------| | "I already know the implementation, writing a test first is wasteful" | The test isn't just for verification — it's a specification. Write it. | | "This is too simple to need TDD" | Simple code is the EASIEST to TDD. No excuse to skip it. | | "I'll write the tests after" | Tests written after code are weaker — they test what you wrote, not what you should have written. | | "The test framework isn't set up yet" | Set it up. That's step 0, not a reason to skip testing. | | "I wrote code first but I can just add a test now" | Delete the code. Write the test. Watch it fail. THEN rewrite the code. This is the discipline. | | "I need to see the implementation to know what to test" | You need to see the REQUIREMENTS to know what to test. Read the spec, not the code. | | "Refactoring isn't needed for this small change" | Small changes accumulate. Refactor now while context is fresh. |

Red Flags (you're doing it wrong if...)

• You wrote more than 10 lines of production code before running a test
• Your test passes on the first run without new code
• You're "refactoring" by adding new features
• You have untested edge cases and you're moving to the next feature
• You wrote multiple tests before writing any production code (write one at a time)

Output

After completion, show:

• How many RED-GREEN-REFACTOR cycles completed
• All passing test names
• Files created/modified
• Any remaining edge cases not yet covered

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: huzaifa525
• Source: huzaifa525/claude-code-optimizer
• License: MIT
• Homepage: https://www.npmjs.com/package/claude-code-optimizer

Install and usage instructions live in the source repository linked above.