Content
96%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.
An exemplary router body: lean, fully concrete, and unambiguous, with correct two-branch progressive disclosure into the playbooks. The only improvement is signposting the shared references/api-reference.md (or the conditions for loading it) so the largest bundle file is discoverable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every line is directive routing guidance ("Route the current request before inspecting files, planning, invoking a specialist, or editing code", "Never load the other playbook later, including after context compaction"); there is no concept explanation or padding, and every token earns its place. | 5 / 5 |
Actionability | Gives exact file paths per branch (references/planner.md, references/iterate.md), an ordered handoff sequence, explicit prohibitions, and concrete decision criteria with tie-breakers ("Observable workspace state wins over wording", "explicit user selection of a child route overrides automatic routing") — fully executable guidance for an instruction-only skill. | 5 / 5 |
Workflow Clarity | The single routing action is unambiguous, sequenced ("State the selected route in one sentence" then "Read exactly one internal playbook immediately"), and backed by edge-case disambiguation rules; no destructive or batch operation is involved, so no validation cap applies. | 5 / 5 |
Progressive Disclosure | The router cleanly points to two well-signaled, one-level-deep playbooks, but the bundle's 60 KB references/api-reference.md is referenced by neither SKILL.md nor either playbook, leaving a large content file without a navigation path — a minor organization gap that keeps it below the top anchor. | 4 / 5 |
Total | 19 / 20 Passed |