Content
77%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 highly actionable with executable CLI examples and a clear single-action workflow for a read-only tool, but it inlines very large JSON schemas that inflate token cost and would be better placed in reference files. Progressive disclosure is the main weakness.
Suggestions
Move the full Input and Output JSON schemas into a references/ file (e.g. references/schema.md) and keep only a short summary plus a link in SKILL.md to reduce inline token load.
If the schemas must stay inline, trim the deeply nested $defs (e.g. UnityEngine.Bounds, SerializedMember) to the fields Claude actually needs and reference the rest.
Add a one-line 'Output fields' summary table at the top of the Output section so the key result shape is visible without parsing the full schema.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The prose is lean and assumes Claude's competence, but ~290 lines of inline output JSON schema constitute a heavy token load that could be tightened or moved to a reference file. | 3 / 5 |
Actionability | Provides copy-paste ready, executable bash commands (unity-mcp-cli run-tool scene-get-data --input '...', plus --input-file and stdin variants) with concrete path-syntax examples and troubleshooting, covering the common cases. | 5 / 5 |
Workflow Clarity | This is a single-purpose read-only tool, so the simple-skill exception applies: the action is unambiguous (choose scene, optionally scope with paths/viewQuery, run the CLI) and a troubleshooting step is included. | 5 / 5 |
Progressive Disclosure | Section headers provide structure, but no bundle/reference files exist and the large input/output schemas are inlined entirely rather than split into a separate reference, which is the kind of content that belongs one level deep. | 3 / 5 |
Total | 16 / 20 Passed |