Content
78%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 exemplary lean skill body: well-organized, no padding, honest scope, and clear sequencing with sensible guardrails. Its one real weakness is actionability — the workflow names tools and categories but never shows an actual command, leaving the executor to reconstruct xcodebuild/swift test invocations on its own.
Suggestions
Add one or two copy-paste-ready command templates under Workflow, e.g. 'xcodebuild test -scheme <scheme> -only-testing:<target>/<class>' and 'swift test --filter <pattern>', so focused execution is executable rather than inferred.
Insert an explicit validation/feedback step, e.g. 'Re-run only the failing test to confirm it reproduces before classifying it as a real failure vs. flake', to close the minor checkpoint gap in the workflow.
For step 1, give the concrete detection cue (e.g. 'Package.swift present → swift test; .xcodeproj/xcworkspace present → xcodebuild test') instead of leaving the detection method unstated.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~25-line body is lean with zero padding: every line is either an argument definition, a workflow step, or a guardrail ('Detect whether the repo uses `xcodebuild test` or `swift test`', 'Avoid rerunning the full suite if a focused rerun is possible'), and no concepts Claude already knows are re-explained. It fully matches 'lean and efficient; assumes Claude's competence; every token earns its place' — the only near-redundancy, the one-line summary repeating the description, is negligible. | 5 / 5 |
Actionability | Guidance is directionally concrete — named tools ('xcodebuild test', 'swift test'), a concrete failure taxonomy ('compile, assertion, crash, env/setup, or flake'), and explicit arguments — but no executable commands or examples are given (no 'xcodebuild test -scheme X -only-testing:...' or 'swift test --filter'), so key execution details are missing. This sits at 'some concrete guidance but incomplete ... missing key details', above the vague-hints level of 2 because the tools, categories, and parameters are specifically named. | 3 / 5 |
Workflow Clarity | The four-step sequence (detect runner → prefer focused execution → classify failures → summarize blocker and next step) is clear and coherent, and the guardrails supply decision checkpoints ('Distinguish build failures from actual failing tests', 'Avoid rerunning the full suite if a focused rerun is possible'), fitting 'clear sequence with most checkpoints present; minor validation gaps'. It is not a 5 because there are no explicit feedback-loop steps (e.g., re-run the focused failing test to confirm flake vs. real failure) and not a 3 because the checkpoints are stated rather than merely implicit, though they are phrased as principles rather than inline validation steps. | 4 / 5 |
Progressive Disclosure | The skill has no bundle files (references/, scripts/, assets/ are absent) and the body is under 50 lines, so the simple-skill exception applies: well-organized sections (Arguments, Workflow, Guardrails) with no content that belongs in a separate file. Per the guideline 'for skills under 50 lines with no need for external references, progressive disclosure can score 5 with just well-organized sections', this earns a 5. | 5 / 5 |
Total | 17 / 20 Passed |