Content
100%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 bootstrap-style body: lean, fully executable, with an explicit output-validation sentinel and recovery path, and clean one-level-deep reference navigation. No weaknesses were identified against the rubric anchors.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is ~25 lines with zero padding and no explanation of concepts Claude already knows; every sentence is a directive (run this command, re-run after upgrade, check for the end-of-guide sentinel, use these two references), so every token earns its place. | 5 / 5 |
Actionability | Fully executable, copy-paste-ready commands are given ('ego-browser skill', plus redirect-to-file variants for both bash and PowerShell), and delegating the per-version usage guide to the CLI's own output is a concrete, complete instruction for the common case. | 5 / 5 |
Workflow Clarity | The sequence is clear: run 'ego-browser skill' before any browser task, follow its output, re-run after upgrade or when instructed, and validate via the explicit '[ego-browser:skill] end of guide' sentinel with a concrete recovery loop (save output to a file and read it in parts) if truncated. The skill is not destructive or batch-oriented, so no workflow cap applies. | 5 / 5 |
Progressive Disclosure | The body is a lean overview with two well-signaled, one-level-deep references, each carrying an explicit read-when condition ('read when ego-browser is not found', 'read before clearing any cookie, cache, or storage'). Both referenced files exist in the bundle and reference no further .md files, and install.md appropriately points to scripts/install.sh, matching the actual bundle structure. | 5 / 5 |
Total | 20 / 20 Passed |