Content
50%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-structured, thorough DDD analysis methodology with concrete patterns, a scoring formula, and output templates, but it is verbose with redundant and textbook content and keeps everything inline rather than splitting reference material into bundle files. Every dimension lands at level 2.
Suggestions
Remove textbook DDD explanations (bounded context definition, ubiquitous language, integration pattern glossary) and the duplicated decision tree in Quick Reference to tighten conciseness.
Add a short worked example with filled-in output (a real domain map and cohesion matrix) so the templates become concrete and actionable rather than placeholder-only.
Extract the output format templates, anti-patterns, and quick reference into reference files under references/ and link to them from SKILL.md, plus add an explicit validation checkpoint (e.g. 're-score cohesion after each split') to the phase workflow.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body explains DDD concepts Claude already knows (bounded context, ubiquitous language, the Shared Kernel/Conformist/Anti-corruption Layer integration patterns) and repeats material—the subdomain decision tree appears in both Phase 3 and Quick Reference—but retains enough original methodology to avoid being pure padding, fitting level 2 rather than level 1. | 2 / 3 |
Actionability | It gives concrete scan patterns (`@Entity`, `*Service`, `*Controller`), a numeric cohesion formula, and copy-paste output templates, but the prescribed actions are advisory ("Suggest separation", "Suggest interface-based integration") with no worked example, leaving it short of fully executable guidance at level 3. | 2 / 3 |
Workflow Clarity | The six phases are clearly sequenced and backed by checklists, but there are no explicit validation/feedback checkpoints within the process—checkpoints are implicit rather than the explicit validate→fix→retry loops required for level 3. | 2 / 3 |
Progressive Disclosure | Sections are well organized, but the skill is a single monolithic ~430-line file with no external references at all, and content that could be split out (output templates, anti-patterns, quick reference) is all inline, matching level 2 rather than the one-level-deep referenced structure of level 3. | 2 / 3 |
Total | 8 / 12 Passed |