Content
86%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 concise, highly actionable instruction skill that maps each product question to a concrete MCP tool call and includes explicit validation gates for navigation and diagnosis. The only minor gap is that workflows are expressed as conditional rules rather than numbered sequences with full error-recovery loops.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence, providing tool-call snippets and tight rules without explaining basic concepts; a small amount of overlap between the '原则' restatement and rule 5 keeps it just below fully lean. | 4 / 5 |
Actionability | It provides copy-paste-ready MCP invocations with exact parameters mapped per question type ('mcp__assistant__product_info({ source: "manifest", section: "routes" })' etc.), fully executable and covering the common cases. | 5 / 5 |
Workflow Clarity | Conditional routing is clear (query the right section by problem; fall back to the compact index when unsure) with an explicit validation checkpoint ('只有导航工具成功后,才能告诉用户点击生成的入口'), but it is rule-based rather than a numbered sequence with full feedback loops. | 4 / 5 |
Progressive Disclosure | At roughly 50 lines with no need for external reference files (data comes from live MCP tools), it is well-organized into clear sections (原则, 当前安装包, 导航与诊断, 信息优先级), fitting the simple-skill exception for well-organized content. | 5 / 5 |
Total | 18 / 20 Passed |