Content
100%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.
An exemplary lean instruction skill: exact commands and flags, an explicit validation step keyed to a named JSON field, defined failure and permission-escalation paths, and zero padding. Nothing is explained that Claude doesn't need, and every operational detail it does need is present.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every token earns its place: the body gives only non-obvious, load-bearing details such as 'Code Mate installs the CN executable qoderclicn, not qoder' and 'published exit-code behavior is incomplete, so a zero exit alone is not success', with no explanation of concepts Claude already knows. | 5 / 5 |
Actionability | Fully executable guidance: a concrete availability check ('command -v qoderclicn'), a copy-paste-ready command ('qoderclicn -p "<prompt>" -o json --no-session-persistence') with an explicit quoting rule, a named success field ('check is_error'), and worked handling for availability and auth failure. No gaps distinguish it from anchor 4. | 5 / 5 |
Workflow Clarity | Clear sequence (set working directory and timeout, check availability, run one task) with an explicit validation checkpoint that distrusts exit codes ('a zero exit alone is not success') and defined stop-and-ask recovery paths for missing tool or authentication. It is a simple, non-destructive skill whose single action is unambiguous, matching the rubric's simple-skill exception for a 5. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines with no need for external references, and its content is well organized into two clear sections ('Run' and 'Authentication And Permissions'), which the rubric explicitly scores as 5 for simple skills. | 5 / 5 |
Total | 20 / 20 Passed |