agentsclimarketplace

Hyper tdd

Skill zeikar/hyperclaude/skills/hyper-tdd

Claude Code plugin: Claude builds, Codex critiques — a gated research → plan → review → ship pipeline, with autonomous multi-agent revise loops that self-converge.

Install
npx -y skills add zeikar/hyperclaude --skill hyper-tdd

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when about to write or modify behavior-bearing code — functions, business logic, bug fixes — before writing any production code. Also when the user invokes /hyperclaude:hyper-tdd. Forces a failing test first so you watch it fail for the right reason.

SKILL.md

1.6 KB, as published. Nobody here has run it

hyper-tdd

Test-driven discipline tailored to this author's preferences.

The loop

  1. Write the failing test first. Name the behavior, not the implementation. The test should compile/import but fail because the behavior doesn't exist yet.
  2. Run it. Confirm it fails for the right reason. "Cannot find function" is fine. "ReferenceError because of typo in test" is not — fix the test.
  3. Write the minimum code to make it pass. Resist scope creep. No "while I'm here" edits.
  4. Run it. Confirm it passes.
  5. Commit. One behavior, one commit. Commit messages describe the behavior, not the code.
  6. Refactor if there's a clear smell. Tests must still pass after.

Skip TDD when

  • Touching only documentation, comments, or formatting.
  • Renaming a symbol (an LSP/grep operation, not a behavior change).
  • One-shot migration scripts that run once and get deleted.
  • Config-only changes (e.g., updating a .gitignore).

What "minimum" means here

Implement only what the current test requires. If you need a helper, add the simplest version. Do not pre-build for tests you might write later — write that test first, then the helper.

What good test names look like

  • ✅ slugify drops non-ASCII characters
  • ✅ cli --dry-run prints JSON without spawning codex
  • ❌ test_slugify (says nothing)
  • ❌ slugify works correctly (vague)

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.