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 well-structured, actionable body that routes by intent, surfaces destructive-operation guards and retry boundaries, and pushes deep flag detail to one-level-deep reference files. Minor conciseness redundancy and prose-form (rather than numbered) workflows keep it just short of top marks.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Lean and assumes Claude's competence (no 'what is Markdown' padding), but the 快速决策 routing bullets (lines 20-24) partly duplicate the shortcuts table and description verbs; not a 5 because of that minor redundancy that could be trimmed. | 4 / 5 |
Actionability | Provides concrete commands (`lark-cli markdown +create/+diff/+fetch/+patch/+overwrite`), specific flags (--folder-token vs --wiki-token, --content/@file/-, --file-token), and a copy-paste-ready BAD/GOOD regex example covering the common escape gotcha; fully executable. | 5 / 5 |
Workflow Clarity | Patch sequence is explicit (download → local replace → overwrite) with an empty-content validation guard and an error-classification feedback loop (stop on missing scope/permission denied/not found/quota vs limited retry on rate_limit/server_error); not a 5 because the sequence is prose rather than a numbered checklist with explicit re-validate loops. | 4 / 5 |
Progressive Disclosure | Clear overview with a shortcuts table linking to five verified one-level-deep reference files (lark-markdown-create/diff/fetch/patch/overwrite.md) plus cross-skill links to lark-shared and lark-drive; content appropriately split and easy to navigate. | 5 / 5 |
Total | 18 / 20 Passed |