Content
92%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 an exemplary lean, single-purpose workflow: concrete templates, explicit validation with a feedback loop, and clear sequencing. The only meaningful gap is the absence of one worked example of a finished rule and the implicit reliance on the format-source skill's own interface for the self-test step.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The ~50-line body is lean and instructional throughout: exact file paths, an exact template, and targeted anti-bias guidance ("Do not bias toward Java", abstract-identifier requirements) with no explanations of concepts Claude already knows. Every token earns its place, matching anchor 5. | 5 / 5 |
Actionability | Guidance is mostly executable: exact paths (`.claude/skills/format-source/SKILL.md`), a copy-paste markdown template, a commit message format, and a concrete self-test procedure. It falls short of anchor 5 because no filled-in example rule is shown and the self-test relies on the reader knowing how to scope `/format-source` to specific files and manual rules only. | 4 / 5 |
Workflow Clarity | Five clearly sequenced steps (inspect, dedupe, write, self-test, commit) with explicit stop-and-ask checkpoints for ambiguity and a full validate-fix-retry feedback loop ("revise the rule and repeat until it does") before committing. This matches anchor 5's explicit validation with error-recovery loops. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines, needs no external reference files (none exist in the bundle), and is cleanly organized into Inputs and a five-step Workflow with headings. Per the simple-skill guideline, well-organized sections alone warrant a 5. | 5 / 5 |
Total | 19 / 20 Passed |