Content
92%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
Excellent body content: a lean, fully actionable TDD workflow with explicit validation checkpoints, a feedback loop, and a done checklist. The single structural flaw is that the bundled reference file writing-good-tests.md is never referenced from the body, so its substantial guidance is unreachable via navigation.
Suggestions
Add a one-line pointer to the reference in the body, e.g. under a '## Writing good tests' heading: 'When writing or changing tests, follow [writing-good-tests.md](references/writing-good-tests.md)' — this makes the existing bundle file discoverable and resolves the only structural gap.
Optionally note how to identify the project's test command (e.g., check package.json scripts, Makefile, or CI config) so the 'Run the project's test command' step is concrete even in unfamiliar repositories.
Consider mentioning the mutation check from the reference (at least one test should fail per realistic mutation) in the 'Before you say it is done' checklist, tying the done criteria to the deeper guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is 23 lean lines with no padding and no explanation of concepts Claude already knows — every directive ("Use a clear name and the real code, not mocks", "Add nothing the test does not need") earns its place. | 5 / 5 |
Actionability | Concrete, directive guidance with explicit error conditions and recovery ("The test must fail because the behavior is missing, not because of a typo or a missing import. If it passes, it tests nothing new: change the test"). As an instruction-only skill whose guidance is fully actionable, the absence of code is not penalized per the rubric's scoring note. | 5 / 5 |
Workflow Clarity | The five-step cycle is clearly sequenced with explicit validation checkpoints ("Watch it fail", "Run the full suite. Every test passes, with no errors or warnings"), a feedback loop for error recovery (a passing new test means change the test), and a completion checklist. | 5 / 5 |
Progressive Disclosure | The body itself is well-organized and lean, but the bundle contains references/writing-good-tests.md (8KB of test-writing guidance) that the body never mentions or links, making the reference undiscoverable from SKILL.md — matching "references present but not clearly signaled". It is not 4 because the sole bundle reference is entirely un-signaled rather than a minor gap, and not 2 because the body is not an inlined monolith. | 3 / 5 |
Total | 18 / 20 Passed |