Content
75%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.
Highly actionable reference content with copy-paste-ready commands, accurate flag semantics, and a stated compilation precondition. The main weakness is redundancy — the Examples section repeats most of the commands already demonstrated in Options — which costs tokens without adding guidance.
Suggestions
Cut the Examples section down to only combinations not already shown inline (e.g., combining --run with --grep, and multi-suite globs), removing the ~6 entries that duplicate the Options section verbatim.
Consolidate the duplicated script invocations so each canonical usage appears exactly once, letting the Options sections serve as the single source of examples.
If the skill grows, move the suite-name catalog (the 15 suite names under --suite) into a references file and link to it, keeping SKILL.md as a lean overview.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient — it documents only repo-specific facts (scripts, flags, suite names, compilation requirement) with no explanation of concepts Claude already knows. However, the "Examples" section duplicates roughly six commands already shown verbatim in the Options section (e.g. "--run src/vs/workbench/services/search/test/browser/search.integrationTest.ts", "--grep \"TextSearchProvider\"", "--suite git", "--suite 'api*'"), which is more than minor trimming. | 3 / 5 |
Actionability | Every command is fully executable and copy-paste ready: exact script invocations, real source paths, glob patterns ("**/search/**/*.integrationTest.js"), the full list of suite names, and flag combinations covering the common cases. No gaps between the guidance and what can actually be run. | 5 / 5 |
Workflow Clarity | The decision logic is clearly sequenced — behavior "without filters", "with --run or --runGlob", "with --grep alone", "with --suite" — and the compilation precondition ("Ensure the VS Code - Build watch task is running or that compilation has completed before running tests") is an explicit checkpoint. It falls short of 5 because there is no explicit sequence framing or error-recovery guidance (e.g., what to check when tests fail to launch), though no destructive/batch validation cap applies. | 4 / 5 |
Progressive Disclosure | The body is well organized into clearly signaled sections (Scripts, Options, Examples, Compilation requirement, Distinction from unit tests) with no bundle files and no nested references, so navigation is easy. It is not a 5 because at ~118 lines it exceeds the self-contained simple-skill threshold and keeps everything inline — the suite catalog and duplicated examples could live in a reference file to slim the SKILL.md. | 4 / 5 |
Total | 16 / 20 Passed |