Content
100%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 instruction-only skill body: a lean router with a prioritized routing table of exact commands and paths, explicit non-goals, environment branching, and hard bounds against hallucination. Nothing is padded, and every instruction is directly executable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~45-line body has no padding and no concepts Claude already knows; every line is a directive or a routing fact ('Prefer its output over inference', 'One surface per question'). The single rationale sentence about prompt budget is load-bearing, matching the 5 anchor ('lean and efficient; every token earns its place'). | 5 / 5 |
Actionability | Fully concrete for an instruction-only skill: exact commands ('/help <command>', '/skills inspect', 'codewhale-tui doctor'), exact file paths ('docs/CONFIGURATION.md', 'docs/KEYBINDINGS.md'), exact tool usage ('read' with 'path: "docs/..."'), and explicit negative guidance ('do not invent one', 'do not invent a flag'). Per the scoring notes, the absence of code is not penalized when guidance is this actionable. | 5 / 5 |
Workflow Clarity | The routing table is explicitly prioritized ('in this order', entries 1–5) with per-entry decision guidance, a clear checkout vs. non-checkout branch, and error-recovery bounds ('the surface wins', 'say so and stop'). No destructive or batch operations apply, and the simple-skill exception holds, matching the 5 anchor's clear sequence with error-recovery loops. | 5 / 5 |
Progressive Disclosure | Under 50 lines with no bundle files needed and well-organized sections; the 'docs/*.md' mentions are routing targets rather than skill references and are each one level deep. Per the rubric guideline, an under-50-line skill with no need for external references scores 5 on well-organized sections alone. | 5 / 5 |
Total | 20 / 20 Passed |