Content
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is concise, well-organized, and gives an executable command with a worked example, but it underspecifies the required environment variables the script depends on, which is the main actionability gap. Otherwise the structure and workflow are clear for a simple skill.
Suggestions
Document the required environment variables (SEARCH_BOT_ID, MODEL_SEARCH_API_KEY/ARK_API_KEY) and how to obtain them from the console, since search_web.py exits without them.
Mention the optional '--stream' flag and 'model=thinking' deep-thinking invocation in the usage steps rather than only listing '深度思考模式' as an advanced feature.
Add a brief note on expected failure behavior (e.g. missing bot ID prints an error) so Claude can diagnose a failed run.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and well-sectioned (~40 lines) with no padding explaining what web search is; every section earns its place and it assumes Claude's competence. | 3 / 3 |
Actionability | Provides a concrete executable command and example ('python scripts/search_web.py "<question>"'), but omits key run-time details the script requires — the SEARCH_BOT_ID and MODEL_SEARCH_API_KEY/ARK_API_KEY environment variables — leaving the guidance incomplete. | 2 / 3 |
Workflow Clarity | This is a simple single-task skill and the single action is unambiguous: prepare a question, cd to the directory, run the script, read the answer; no destructive/batch steps require validation checkpoints. | 3 / 3 |
Progressive Disclosure | Well-organized into clear sections with the only bundle artifact being the real, correctly-referenced scripts/search_web.py; for a sub-50-line single-purpose skill this clean structure satisfies the rubric. | 3 / 3 |
Total | 11 / 12 Passed |