Content
82%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, well-structured, mostly actionable skill doc: the command and example are executable and the output contract is clearly specified. The main weaknesses are undocumented prerequisites (SEARCH_BOT_ID/API key env vars), vague setup instructions ('cd到对应的目录'), and advanced features (model=thinking, console toggles) with no explanation of how to actually enable them.
Suggestions
Document required environment variables up front (e.g., '运行前需设置 SEARCH_BOT_ID 和 MODEL_SEARCH_API_KEY/ARK_API_KEY'), since the script exits without them — the most common failure is currently undiscoverable.
Make setup concrete: state the exact directory (skill root) instead of 'cd到对应的目录', and show how model=thinking is actually set (command flag, env var, or payload field) in the 示例.
Add a brief error-recovery note (e.g., 'SEARCH_BOT_ID 未设置时会提示；请在控制台获取后重试') so the error path has a fix-and-retry loop instead of just 'prints error info'.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is ~40 lean lines of pure operational content (sections: 适用场景, 使用步骤, 输出格式, 示例, 高级功能) with zero padding and no explanation of concepts Claude already knows, matching 'every token earns its place'. It is not merely a 4 because there is no over-explanation anywhere to trim. | 5 / 5 |
Actionability | The core command `python scripts/search_web.py "<question>"` plus a runnable example is concrete, matching 'concrete code or commands with minor gaps'. The gaps: the script requires `SEARCH_BOT_ID`/API-key env vars (verified in scripts/search_web.py) that are never documented, '运行之前cd到对应的目录' is vague about which directory, and '通过设置 model=thinking 开启' gives no actual mechanism. These are gaps, not absence of guidance, so not a 3. | 4 / 5 |
Workflow Clarity | The 3-step sequence (prepare question → run script → read answer incl. error output) is clear and unambiguous for a simple skill, matching 'clear sequence with most checkpoints present; minor validation gaps'. Not a 5 because error handling stops at '若调用失败，将打印错误信息' with no fix-and-retry guidance (e.g., missing SEARCH_BOT_ID is the most likely failure but unmentioned). | 4 / 5 |
Progressive Disclosure | The skill is under 50 lines with well-organized sections and a real, existing referenced script (scripts/search_web.py verified), so structure is good — matching 'good structure; most content appropriately placed; minor organization gaps'. Not a 5 because advanced options (thinking mode mechanism, console toggles, stream flag) have no documented path — neither in SKILL.md nor via a reference — leaving those features undiscoverable. | 4 / 5 |
Total | 17 / 20 Passed |