Content
96%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.
An exceptionally lean, expert-level reference: executable examples, exact commands, and high-value gotchas with explicit decision rules for every feature. The only structural note is that the body is self-contained at ~100 lines with no reference files, which is defensible but leaves minor room to split detail out.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every section carries non-obvious Ginkgo semantics or gotchas ("BeforeAll/AfterAll are legal only inside an Ordered container", "--repeat=N runs 1+N times", "not compatible with go test --count") with zero padding or explanation of concepts Claude already knows. | 5 / 5 |
Actionability | Copy-paste-ready Go snippets (Ordered container with BeforeAll/DeferCleanup/DeferOnFailure, OncePerOrdered BeforeEach) and exact CLI invocations ("ginkgo --until-it-fails -p --randomize-all", "--repeat=N", "--flake-attempts=3") cover the common cases. | 5 / 5 |
Workflow Clarity | Decision guidance is unambiguous ("Prefer one It + By by default", "Default to none — debug the flake", "use Ordered only when needed"), and each tool's semantics, failure behavior, and anti-patterns are stated explicitly. No destructive or batch operations require validation loops. | 5 / 5 |
Progressive Disclosure | Well-organized sections that delegate depth to sibling skills ("The full decorator signatures → ginkgo:decorators"), but the ~100-line body has no bundle references and inlines detail (e.g. the FlakeAttempts × Ordered rerun interaction) that could live in a reference file. | 4 / 5 |
Total | 19 / 20 Passed |