Content
71%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.
A well-structured overview skill: clear coverage list, a gated workflow with compile-before/verify-after checkpoints, and proper progressive disclosure to a real, verified reference file. The main weakness is redundancy — the Constraints section repeats one rule four times and the 'When to use' section restates the description — plus a missing recovery step when post-change verification fails.
Suggestions
Consolidate the Constraints section to three bullets — compile before applying, stop and escalate to the user if compilation fails, run `clean verify` after — removing the PREREQUISITE/BLOCKING CONDITION restatements of the same rule.
Drop the 'When to use this skill' section: it duplicates the frontmatter description's trigger phrases verbatim and adds no body-specific information.
Add a short failure path to workflow step 4 (e.g., 'If `clean verify` fails, fix and re-run before reporting') to close the feedback loop after applying changes.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lean and assumes Claude's Quarkus/JUnit knowledge, but the Constraints section states the same compile rule four ways ("MANDATORY: Run ./mvnw compile", "PREREQUISITE: Project must compile", "SAFETY: If compilation fails, stop immediately", "BLOCKING CONDITION: Compilation errors must be resolved by the user") and "When to use this skill" duplicates the description's trigger phrases verbatim. This is 'mostly efficient but includes some unnecessary explanation or could be tightened' rather than the noticeably padded anchor 2, since the redundancy is confined to two short sections. | 3 / 5 |
Actionability | Gives concrete, executable commands ("Run `./mvnw compile` or `mvn compile`", "Run `./mvnw clean verify` or `mvn clean verify`") and names specific mechanisms (@InjectMock, @InjectSpy, QuarkusTestProfile, *Test/*IT conventions), deferring code patterns to the verified reference file. Not a 5 because the body itself contains no code examples for the common cases — everything pattern-level lives in the reference — and not a 3 because the commands given are directly runnable and the coverage list is concrete. | 4 / 5 |
Workflow Clarity | The four-step workflow (read reference and assess context → gather scope → apply changes → run verification and report) is clearly sequenced, with an explicit compile-before gate ("MANDATORY: Run ./mvnw compile... before applying any change") and a verify-after step. Not a 5 because there is no feedback loop for verification failure — the workflow says "summarize what changed, what was verified" but not what to do if `clean verify` fails — leaving a minor validation gap at anchor 4. | 4 / 5 |
Progressive Disclosure | The body is a concise overview with a single clearly-signaled, one-level-deep reference ([references/421-frameworks-quarkus-testing-unit-tests.md](references/421-frameworks-quarkus-testing-unit-tests.md)), which exists and holds the detailed rules and examples (verified on disk, 661 lines). The reference is surfaced three times (workflow step 1, "BEFORE APPLYING", and the Reference section) and no detail content that belongs in the reference is inlined. This matches the anchor for a clear overview with well-signaled, one-level-deep references. | 5 / 5 |
Total | 16 / 20 Passed |