Content
67%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-designed router/index skill: symptom-to-file routing, cross-skill conflict resolution, and anti-rationalization content are all genuinely useful and unlikely to be things Claude already knows. Its weaknesses are redundant repetition of the same routing map across four sections, and references to skills/*.md files that are not present alongside the skill, making the disclosure structure unverifiable.
Suggestions
Collapse the duplicated routing into one canonical form: keep the Quick Reference table and either the DOT tree or the numbered list, and cut the Critical Patterns and Example Invocations sections that restate the same file mappings.
Verify the skills/*.md bundle actually ships with the skill (no skills/ directory is present here); broken reference targets would make the entire router non-functional, so confirm the paths or use bundle-relative paths that resolve.
Keep at most one or two genuinely distinctive routing facts (e.g. the geofencing two-file sequence and the battery-drain handoff) as illustrative examples rather than an eight-row Example Invocations section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Individual lines are terse and there is no over-explanation of known concepts, but the same routing information is presented three to four times: the Quick Reference table, the DOT decision tree, the numbered decision list, the Critical Patterns section, and Example Invocations all repeat the symptom-to-file mapping. That matches anchor 3 (mostly efficient but could be tightened) rather than anchor 4, since whole duplicate sections could be cut. | 3 / 5 |
Actionability | As a router skill, its guidance is concrete and directive: every symptom maps to an exact file ("Location updates never arrive → skills/core-location-diag.md"), conflict rules give explicit ordering ("Try location FIRST"), and the anti-rationalization table cites specifics ("30-60% denial rate"). It fits anchor 4 (mostly executable guidance, minor gaps) rather than anchor 5 because it contains no copy-paste commands/snippets and the referenced files cannot be verified from the skill directory. | 4 / 5 |
Workflow Clarity | The routing sequence is clear and unambiguous: domain selection (Core Location vs MapKit) then task type (pattern vs API syntax vs diagnostics), with numbered fallback steps and explicit cross-skill ordering in Conflict Resolution. It is not a destructive or batch workflow, so the validation cap does not apply; it stays at anchor 4 rather than 5 because the four parallel routing presentations (table, DOT graph, numbered list, examples) leave it unclear which one is authoritative. | 4 / 5 |
Progressive Disclosure | The body is a pure overview that consistently delegates detail to one-level-deep, clearly signaled references ("See skills/core-location.md", "See [skills/mapkit-ref.md]"), which matches the top anchors. However, none of the referenced skills/*.md files exist in this skill's directory (no references/, scripts/, or skills/ bundle is present), so the navigation structure cannot be verified against an actual bundle, holding it at anchor 4 rather than 5. | 4 / 5 |
Total | 15 / 20 Passed |