Content
71%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.
Well-structured CLI skill with strong progressive disclosure and actionable, validated discovery workflows. The main weakness is redundancy across the search-disambiguation tips, which inflates the body without adding new guidance.
Suggestions
Collapse the repeated search-vs-list disambiguation blockquotes (任务搜索技巧, 任务清单搜索技巧, 意图区分补充) into a single concise decision rule to reduce redundancy.
Add an explicit validation/confirmation checkpoint for destructive operations (tasks.delete, tasklists.delete) to lift workflow_clarity above the destructive-cap cap.
Replace generic '<resource>'/'<method>' placeholders in the API Resources section with one fully concrete end-to-end example command to close the actionability gap.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly dense and non-padded, but the search-vs-list disambiguation is re-explained across several consecutive blockquotes (任务搜索技巧, 任务清单搜索技巧, 意图区分补充) that restate the same logic and could be tightened. | 3 / 5 |
Actionability | Provides concrete executable commands — 'lark-cli task --help', 'lark-cli schema task.<resource>.<method>', 'lark-cli task tasklists tasks --params \'{...}\' --as user' — plus a shortcut table; minor gaps where templates use '<resource>'/'<method>' placeholders. | 4 / 5 |
Workflow Clarity | The numbered command-discovery flow has an explicit feedback loop (step 5 returns to step 2 on unknown_subcommand) and validation (run schema before API calls); destructive operations are only listed as methods without their own validation checkpoint. | 4 / 5 |
Progressive Disclosure | Clear overview with a well-signaled one-level-deep shortcut table linking to 17 real reference files, plus separated API Resources and permission tables — all 17 referenced files exist and navigation is easy. | 5 / 5 |
Total | 16 / 20 Passed |