Content
81%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 content is a strong, executable runbook: concrete commands, explicit input validation, error-recovery guidance, and a clear four-step workflow. Its weaknesses are moderate redundancy (repeated flag explanations and near-verbatim Examples) and an inline-heavy structure. Trimming the duplicated material would improve token efficiency without losing actionability.
Suggestions
Remove the duplicated flag explanations under 'Build and run the command' — the two command variants and the Usage line already convey when --base and --panel apply.
Cut or compress the Examples section, which restates the Instructions nearly verbatim; a single short example with realistic output framing would suffice.
Consider moving the Panels detail into a short reference file (or tightening it to 3-4 lines) so the core workflow stays scannable.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient — it defines the domain term and gives operational guidance Claude would not infer — but there is noticeable repetition: the usage line is restated in Instructions, the --base/--panel bullet explanations restate what the two command variants already show, and the Examples section replays the instructions nearly verbatim. This fits 'mostly efficient but includes some unnecessary explanation or could be tightened'; it is not a 4 because the redundancy spans several sections. | 3 / 5 |
Actionability | Commands are copy-paste ready and cover the common cases: "roborev review --from-skill --branch --wait --type lookahead" for the default case, and a complete base-branch snippet with "git rev-parse --verify --end-of-options" plus a heredoc template. Error-remediation commands ("roborev daemon status", "roborev init", "/roborev-fix <job_id>") and concrete job-ID extraction guidance ("Look for it in the `Enqueued job <id> for ...` line") are also specified. | 5 / 5 |
Workflow Clarity | The four steps (validate inputs → build and run command → present results → offer next steps) are clearly sequenced, with an explicit validation checkpoint ("git rev-parse --verify --end-of-options ... || exit 1" and "If validation fails, inform the user the ref is invalid. Do not proceed.") and error-recovery feedback loops (report daemon/repo errors and suggest "roborev daemon status", "roborev init", or re-running). The destructive/batch cap does not apply since the skill only runs a read-only review. | 5 / 5 |
Progressive Disclosure | The body is well-organized with clear section headers (Usage, Sandbox access, When NOT to invoke, Instructions, Examples, See also) and all content lives in one appropriately sized file with no nested references. However, the skill exceeds 50 lines and detailed material (the Panels subsection and the largely duplicative Examples) is inlined where a leaner core or a single reference file could carry it, so it sits at 'good structure; minor organization gaps' rather than the clear split-and-reference ideal. | 4 / 5 |
Total | 17 / 20 Passed |