Content
53%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 appropriately brief for a simple skill but skeletal: its single usage command references a script that does not exist in the bundle and uses a mismatched interpreter, making the guidance unreliable. Fixing the command and removing the redundant overview would raise quality substantially.
Suggestions
Fix the usage command — either ship an actual read.sh in scripts/ and invoke it with bash, or use a correct interpreter — so the example is copy-paste executable.
Remove the 概述 section that duplicates the frontmatter description verbatim, or replace it with genuinely useful notes (output format, error handling for missing/unreadable files).
Correct the parameter table: describe <file> as the file path to read, not 'the text to read', and document expected output or exit behavior.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean with no over-explanation of concepts Claude already knows, but the 概述 section duplicates the frontmatter description verbatim, so not every token earns its place. | 4 / 5 |
Actionability | It gives one concrete command ('python read.sh "<file>"'), but that command is not executable: no read.sh exists in the bundle (no scripts/ directory) and invoking a .sh file with python is an interpreter mismatch, leaving key execution details broken or missing. | 2 / 5 |
Workflow Clarity | As a single-action skill the one step is stated, but it is ambiguous — the command is questionable, the parameter table describes <file> as '要读的文本' (the text to read) rather than a path, and there is no guidance on output or error cases. | 3 / 5 |
Progressive Disclosure | The skill is under 50 lines with well-organized sections (概述, 基本使用方法, 命令行参数) and needs no external references, but it dangles a reference to a nonexistent script (read.sh), a minor organization gap. | 4 / 5 |
Total | 13 / 20 Passed |