Content
80%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, actionable skill body — every token earns its place and commands are concrete — held back only by missing validation around its destructive/batch operations. Progressive disclosure is appropriately trivial for a skill this size.
Suggestions
Add an explicit verification step for the destructive and batch operations, e.g. checking the osascript exit status before reporting "trash: emptied" and handling failed pulls — missing validation in destructive/batch workflows caps workflow_clarity at 3.
Note the shell requirement (zsh) for `dir=${repo:h}` or use portable syntax like `dir=$(dirname "$repo")` so the loop runs in any shell.
Make the skip-dirty check executable, e.g. `if [ -z "$(git -C "$dir" status --porcelain)" ]; then git -C "$dir" pull --ff-only; else echo "skipped: $dir"; fi`, so the guardrail is not left to per-repo interpretation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every line instructs: a one-line trigger, three concrete command blocks, one guardrail sentence, and a terse reporting format. No explanation of concepts Claude already knows, no padding. Matches the lean anchor exactly. | 5 / 5 |
Actionability | Commands are concrete and nearly copy-paste ready, but there are minor gaps: `dir=${repo:h}` is zsh-only with no shell stated, and the loop prints `git status` without acting on it, leaving the skip-dirty decision implicit per repo. Not 5 because of these executable-edge gaps; not 3 because the guidance is real, runnable code, not pseudocode. | 4 / 5 |
Workflow Clarity | The four-step sequence is clear and includes safety-conscious touches (`--ff-only`, "Skip dirty repos... Report skipped paths"), but this is a batch (repo loop) plus destructive (emptying Trash) workflow with no explicit validation checkpoints or error-recovery loop — the rubric caps such workflows at 3. Not 2 because the sequence is well defined with some implicit checkpoints (status output, final counts). | 3 / 5 |
Progressive Disclosure | The skill is under 50 lines, has no bundle files, and needs none; content is organized into a clear trigger line plus a well-structured numbered Run section. The rubric's simple-skill exception applies: well-organized sections score 5 without external references. | 5 / 5 |
Total | 17 / 20 Passed |