Content
52%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The skill has a solid structure with a clear validation-first workflow and appropriate delegation to a reference file, but suffers from significant redundancy — the same concepts (multi-module discovery, untrusted POM input, cross-module checks) are repeated verbatim across multiple sections. Actionability is weakened by the complete absence of concrete examples or executable snippets in the main skill file, with all substance deferred to the reference.
Suggestions
Consolidate the repeated multi-module, untrusted-input, and cross-module rules into a single authoritative location (Constraints section) and remove duplicates from the intro bullets and workflow steps.
Add at least one concrete before/after POM snippet or an example xmllint/xpath command to make the skill actionable without requiring the reference file.
Add a post-modification re-validation step (e.g., 'After applying changes, run `./mvnw validate` again to confirm no regressions') to close the feedback loop.
Remove the 'When to use this skill' section since it duplicates the YAML description and adds no new information.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Significant redundancy throughout: the multi-module discovery, cross-module scope, and untrusted POM input rules are each stated 2-3 times (in the intro bullets, Constraints section, and Workflow). The 'What is covered' section and 'When to use this skill' section largely duplicate the YAML description. Much of the constraint text could be consolidated. | 2 / 5 |
Actionability | The workflow provides a clear sequence (validate, parse, expand, recommend) and specifies concrete commands (`./mvnw validate`), but lacks executable code examples for XML querying, actual POM transformation patterns, or concrete before/after snippets. The real actionable content is deferred entirely to the reference file. | 3 / 5 |
Workflow Clarity | The 4-step workflow is clearly sequenced with an explicit validation gate (stop if validation fails) and a safety constraint. However, there's no explicit re-validation step after applying changes, which is a minor gap for a destructive/modification workflow. The feedback loop for post-modification validation is missing. | 4 / 5 |
Progressive Disclosure | The skill appropriately defers detailed examples and patterns to a reference file (`references/110-java-maven-best-practices.md`) with a clear one-level-deep reference. However, since no bundle files were provided, we cannot verify the reference exists or is well-structured. The SKILL.md itself is somewhat long for an overview that delegates to a reference. | 4 / 5 |
Total | 13 / 20 Passed |