Content
88%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 executable, well-sequenced runbook: exact commands for both invocation forms, up-front ref validation with failure handling, explicit error-recovery paths, and conditional presentation and next steps keyed to the verdict. It stays lean and assumes competence, with only minor tightening or splitting opportunities (Panels detail, duplicated invocation rules).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body stays operational ("The `--wait` flag blocks until the review completes", "These instructions are guidelines, not a rigid script") and teaches nothing Claude already knows; the only trimmable redundancy is that the invocation gate is restated across "Explicit invocation only" and "When NOT to invoke this skill", so it matches "Efficient; minor instances of over-explanation that could be trimmed" rather than the fully lean anchor 5. | 4 / 5 |
Actionability | Fully executable, copy-paste-ready commands cover both cases — `roborev review --from-skill --wait --type design` and the heredoc plus `git rev-parse --verify --end-of-options "$commit^{commit}"` validation snippet — with concrete error remedies (`roborev daemon status`, `roborev init`) and worked examples including job IDs (`/roborev-fix 1042`), matching anchor 5. | 5 / 5 |
Workflow Clarity | Four clearly sequenced steps with explicit validation ("git rev-parse --verify ... || exit 1", "If validation fails, inform the user the ref is invalid. Do not proceed."), feedback loops for error recovery (sandbox escalation retry, daemon/repo failure remedies), and conditional next steps keyed to the verdict (offer `/roborev-fix` only on Fail) — matching anchor 5's "explicit validation steps; feedback loops for error recovery"; the operation is a review, not destructive/batch, so no cap applies. | 5 / 5 |
Progressive Disclosure | Sections are well organized (Usage, Sandbox access, When NOT to invoke, numbered Instructions, Examples, See also) with no bundle files present and nothing that clearly belongs in a separate file, but the ~10-line Panels subsection and the sandbox troubleshooting detail are niche enough to live in a reference file, matching "Good structure; most content is appropriately placed ... minor organization gaps". | 4 / 5 |
Total | 18 / 20 Passed |