Content
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A well-structured router skill: actionable references, a clear decision tree, and good progressive disclosure. Its main weakness is redundancy — the routing table, routing sections, decision tree, simplified list, and example invocations all restate the same dispatch information, inflating the token budget.
Suggestions
Collapse the redundant routing representations: pick either the Quick Reference table OR the Routing Logic sections as the primary dispatch, and drop or sharply trim the others to cut token cost.
Trim the Example Invocations section — it largely duplicates the Quick Reference table and Simplified decision tree; keep only a handful of representative cases.
Consider merging the graphviz Decision Tree and the Simplified numbered list into one canonical routing sequence to avoid maintaining two parallel decision artifacts.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The same routing decisions are restated four to five times — Quick Reference table, 11 Routing Logic sections, a graphviz Decision Tree, a Simplified numbered list, and ~30 Example Invocations — so the body is mostly efficient but carries substantial redundant content that could be tightened. | 2 / 3 |
Actionability | Gives concrete, copy-paste-ready direction throughout — specific file paths ('See skills/app-store-diag.md'), agent launches ('Launch security-privacy-scanner agent'), and slash commands ('/axiom:triage sentry') — which is appropriately actionable for a routing skill. | 3 / 3 |
Workflow Clarity | Provides an explicit, branching decision tree plus a simplified numbered sequence that walks Claude from symptom to the correct sub-skill/agent, with clear routing branches. | 3 / 3 |
Progressive Disclosure | Acts as a lean overview that signals one-level-deep references (specific skills/*.md files and named agents) via a Quick Reference table and per-route Reference lines, with content appropriately split into sub-skills rather than inlined. | 3 / 3 |
Total | 11 / 12 Passed |