Content
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a well-sequenced, actionable workflow with explicit validation and error handling; its main weakness is redundant duplication of the command blocks across steps 2 and 3.
Suggestions
De-duplicate the bash command blocks: define the command once (e.g. in step 2) and reference it from step 3, or factor the base-branch validation snippet into a single labeled snippet reused by both steps.
Avoid restating the full command syntax in the Usage section, step 2, and step 3; show it once and refer back to it.
Consider moving the panels sub-section to a reference file if the skill grows, to keep the main flow lean.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient, but the full bash command block (including the read -r branch <<'ROBOREV_REF' ... git rev-parse validation snippet) is duplicated nearly verbatim in both step 2 and step 3, and the command syntax recurs across Usage, step 2, and step 3, adding redundant tokens. | 2 / 3 |
Actionability | It provides fully executable commands (roborev review --branch --wait with real flags, the git rev-parse --verify validation snippet) and concrete tool guidance (Task tool with run_in_background: true, subagent_type: "Bash"), copy-paste ready. | 3 / 3 |
Workflow Clarity | Five clearly sequenced steps include explicit validation checkpoints (git rev-parse --verify ... || exit 1, "If validation fails... Do not proceed") and an error-recovery branch in step 4, satisfying the feedback-loop requirement. | 3 / 3 |
Progressive Disclosure | No bundle files exist; the self-contained single-task skill is organized into clear sections (Usage, When NOT to invoke, numbered Instructions, Examples, See also) with one-level, well-signaled cross-references to sibling skills and no nested references. | 3 / 3 |
Total | 11 / 12 Passed |