Content
92%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.
The body is an exemplary lean skill: executable code per subcommand, a crisp disambiguation table, and prose reserved for non-obvious pitfalls. The only weakness is the versions section's placeholder variables, which keep the last code block from being directly runnable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and code-first; the only prose is genuinely non-obvious domain gotchas ('load_from_file does not expand ~', 'a list of dicts (unhashable), so compare the id fields'), exactly what Claude could not infer. Every token earns its place. | 5 / 5 |
Actionability | The graph-at and node-history blocks are fully executable with imports and gotcha workarounds, but the versions block uses undefined placeholders ('ontology', 'base_ontology', 'target_ontology') so it is not copy-paste ready. Not 5 for that minor gap; well above 3 since real executable code dominates. | 4 / 5 |
Workflow Clarity | Each subcommand is an unambiguous single read-only action, and the opening table explicitly disambiguates which tool answers which question. No destructive or batch operations are involved, so the simple-skill exception applies and no validation checkpoints are required. | 5 / 5 |
Progressive Disclosure | No bundle files exist, and nothing inlined here belongs in a separate file: the body is short, organized into one clear section per subcommand, and delegates richer temporal work to /semantica:temporal instead of duplicating it. Well-organized sections fully satisfy the no-external-references case. | 5 / 5 |
Total | 19 / 20 Passed |