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 a tight, executable runbook: exact commands with an explicit ref-validation gate, concrete error-recovery guidance, precise result-presentation requirements, and worked examples for both invocation variants. Its only flaws are minor redundancy in restating flag conditions and a length that is borderline for a single-file skill with no reference split.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is operational throughout — it never explains concepts Claude already knows, and the one definitional sentence (look-ahead/peekahead/future leakage) is domain-specific and earns its place. Minor trimmable redundancy keeps it from 5: the usage string appears twice (## Usage and the Instructions lead-in), and the two bullets after the base-branch snippet restate the if/else conditions already given. | 4 / 5 |
Actionability | Guidance is fully executable: exact copy-paste roborev commands for both the no-base and base cases, a heredoc + `git rev-parse --verify --end-of-options` validation snippet, concrete error-recovery commands (`roborev daemon status`, `roborev init`), and where to find the job ID ("Enqueued job <id> for ..." line). Two worked examples cover the common cases; this matches the anchor for copy-paste ready commands with specific examples. | 5 / 5 |
Workflow Clarity | A clear four-step sequence (validate inputs → build and run → present results → offer next steps) with an explicit validation gate ("If validation fails... Do not proceed") and explicit error-recovery branches for daemon/repo failures including re-running the review. This is a read-only review skill, so the destructive/batch cap does not apply, and the validation-and-retry structure matches the top anchor. | 5 / 5 |
Progressive Disclosure | No bundle files exist and none are needed; the single file is organized into clearly signaled sections (Usage, Instructions, Examples, See also) with cross-skill navigation in "See also". It scores 4 rather than 5 because at ~135 lines it exceeds the under-50-line simple-skill exception, and the Panels subsection plus the Examples dialogues are content that could arguably live in a reference file if the skill grows. | 4 / 5 |
Total | 18 / 20 Passed |