Content
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill functions primarily as a table of contents pointing to a reference file, with almost no actionable content in the SKILL.md itself. While the workflow structure and constraints are reasonable, the complete absence of concrete code examples for any of the advertised testing patterns (Mockito, @InjectMock, @InjectSpy, @ParameterizedTest, REST Assured) severely limits its usefulness. The skill would benefit greatly from including at least a minimal executable example for the most common pattern (pure Mockito test vs @QuarkusTest with @InjectMock).
Suggestions
Add at least two concrete, executable code examples inline: one for a pure Mockito test (@ExtendWith(MockitoExtension.class)) and one for @QuarkusTest with @InjectMock, showing the key difference and when to choose each.
Remove the duplicated 'What is covered' bullet list since it repeats the description and 'When to use' sections — replace it with a quick-start code snippet that demonstrates the primary recommended pattern.
Consolidate the constraints section — the five bullet points about compilation can be reduced to two: 'Run ./mvnw compile before changes; run ./mvnw clean verify after. Stop if compilation fails.'
Make workflow step 3 specific: instead of 'Apply framework-aligned changes', provide concrete sub-steps like 'Replace @QuarkusTest with @ExtendWith(MockitoExtension.class) where no CDI wiring is needed' with a before/after code snippet.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The 'What is covered' section largely duplicates the description metadata and the 'When to use this skill' section repeats trigger phrases from the description. The constraints section is somewhat repetitive (multiple bullet points saying essentially 'compile first, stop if it fails'). However, it's not egregiously verbose and doesn't explain concepts Claude already knows. | 2 / 3 |
Actionability | The skill contains zero executable code examples, no concrete commands beyond generic mvnw compile/verify, and delegates all actual guidance to a reference file. There are no copy-paste ready patterns for @InjectMock, @InjectSpy, @ParameterizedTest, or any of the advertised topics. Claude would need to read the reference file to do anything useful. | 1 / 3 |
Workflow Clarity | The workflow has a clear 4-step sequence with compilation prerequisite and verification at the end, which is good. However, the steps are generic ('apply framework-aligned changes', 'gather scope') rather than specific to Quarkus testing. The validation checkpoints (compile before, verify after) are present but the error recovery is limited to 'stop immediately' without guidance on common failure modes. | 2 / 3 |
Progressive Disclosure | The skill correctly references a single external file for detailed guidance, which is good one-level-deep disclosure. However, since no bundle files were provided, we cannot verify the reference exists or is well-structured. The SKILL.md itself is essentially a thin wrapper that delegates almost everything to the reference, making it too thin — the quick-start content that should be inline (at least one concrete example) is missing entirely. | 2 / 3 |
Total | 7 / 12 Passed |