Our review
The skill describes a red, green, refactor loop for features and bug fixes. It emphasizes watching the new test fail for the expected reason before writing production code, then making the smallest change that passes and rerunning the suite. An accompanying note discusses how to write tests that express intent rather than mirror implementation details.
Use it where behavior can be stated as a meaningful automated check. Skip the literal cycle for a purely visual adjustment, exploratory prototype, or work where an existing test already captures the exact defect. The skill is intentionally strict; applying it mechanically can create low-value tests. Choose assertions that protect user-visible behavior and keep the test suite maintainable.
The folder has SKILL.md and one guidance note, with no bundled executable script or direct network call. It tells the agent to run the project’s test commands, so their dependencies and side effects belong to that project. Read the current instructions and decide which test level can actually expose the failure.
For installation, open the linked maintainer folder and read its current SKILL.md, license, and supporting files. Copy the complete folder to a skills location your agent documents, or use an installer your agent supports. Begin with a small, low-risk task and confirm that the agent loaded the intended version. Our install guide covers those checks.
Source and license
Maintainer: Jesse Vincent / obra. Checked: October 18, 2026. The maintainer controls this repository; the current files may differ from the version we reviewed.