Content
78%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.
The body is a well-structured, highly actionable instruction set: concrete search queries, mermaid rendering rules with working examples, a required chapter list, output conventions, and edge-case handling. Its main weaknesses are the absence of validation checkpoints in the workflow and a fully monolithic layout where mermaid rules and the report template could be offloaded to reference files to slim the always-loaded context.
Suggestions
Move the mermaid Chinese-rendering rules and the report-opening template into references/ (e.g., references/mermaid-zh.md, references/report-template.md) and link to them one level deep, keeping SKILL.md as a lean overview.
Add an explicit validation step before output (Step 5), e.g., 'before saving, verify every mermaid block follows the quoting/date rules above and the report meets the minimum section list and 4000-character floor; if not, fix and re-check'.
Trim advisory prose in the 四象限画像 tables and the duplicated chapter requirements in the template example to reduce token cost without losing guidance.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is largely lean and assumes competence — no explanations of what SWOT, MVP, or gantt charts are; it directly gives search query templates, dimension tables, and rendering rules. Score 4 rather than 5 because a few sections could be tightened (e.g., the report-opening template partially repeats chapter requirements, and the 四象限画像 tables carry some advisory prose that could be trimmed). Not 2-3: there is no padding that explains concepts Claude already knows. | 4 / 5 |
Actionability | Fully executable guidance throughout: ready-to-use web search query strings ('"[企业名] 营收 利润 2023 2024"'), complete copy-pasteable mermaid examples for quadrantChart/gantt/timeline/flowchart, an explicit file-naming format ('[企业名]-未来规划建议书-[YYYY-MM].md'), a concrete report skeleton, and concrete quality gates (✅/❌ rules). Covers the common cases directly. | 5 / 5 |
Workflow Clarity | A clear 5-phase pipeline is stated upfront ('1. 信息采集 → 2. 企业画像构建 → ... → 5. 输出报告') and each step is elaborated, with an explicit fallback path for insufficient information ('信息严重不足时... 基于行业推演并明确标注'). Score 4 rather than 5: there are no validation checkpoints (e.g., verifying mermaid renders or that the 4000-character/section requirements are met before saving), though the skill is not destructive/batch so the hard cap does not apply. | 4 / 5 |
Progressive Disclosure | The skill is a single ~275-line monolithic file with no bundle files (no references/, scripts/, assets/ exist). Structure is good (clear numbered sections), but content that would naturally live in reference files — the mermaid Chinese-rendering rules, the report template, the chapter checklist — is inlined. Fits anchor 3 ('some structure; content that should be separate is inline'); not 4 because there is no file split at all, and not 2 because the inline content is well-organized and signposted rather than a wall of text. | 3 / 5 |
Total | 16 / 20 Passed |