TDD Guard Hits 2K Stars Blocking Claude Code From Skipping Tests
TDD Guard hooks into Claude Code to block implementation without failing tests, keeping the coding agent honest across nine test frameworks.
PRO- TDD Guard blocks Claude Code from writing implementation without a failing test first.
- Installs as a Claude Code plugin via three commands, no manual hook editing required.
- Supports Vitest, Jest, Storybook, pytest, PHPUnit, Go, Rust, RSpec, and Minitest.
- Uses PreToolUse, UserPromptSubmit and SessionStart hooks to intercept Write, Edit and TodoWrite calls.
- MIT licensed, 2,000+ stars, requires Node.js 22 or newer to run.
- Evolving into Probity, extending enforcement to Codex and GitHub Copilot CLI.
Anyone who has watched Claude Code power through a feature has probably also watched it invent tests after the fact, silently skip a red step, or dump a full implementation before a single assertion runs. TDD Guard is a Claude Code plugin that refuses to let that happen. It hooks into the agent's tool calls, blocks writes that violate test-driven development, and tells the model what it should do instead.
The project has crossed 2,000 stars on GitHub and is published to npm under the MIT license. Installation is a three-command sequence inside Claude Code that adds the marketplace, installs the plugin, and runs a setup skill that wires up the test reporter for your project.
How the guardrails actually fire
TDD Guard plugs into Claude Code's hook system rather than the model itself, so enforcement is deterministic and happens before any file gets touched. Under the hood it registers three hooks in your Claude settings:
PreToolUsematchingWrite|Edit|MultiEdit|TodoWrite, which intercepts every attempt to modify code or the task listUserPromptSubmit
This story is for Pro members
You've reached the end of the free preview. Upgrade to AlphaSignal Pro to read the full article - and everything else behind the paywall.
Comments
