Content
67%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 well-organized, actionable catalog with copy-paste-ready commands and useful fallback guidance for rate limits. Its main weaknesses are token padding from redundant command examples and generic best-practices advice, and the absence of an explicit step-by-step search workflow with a preferred-source order.
Suggestions
Collapse the three redundant 'By Functionality'/'By Provider'/'By Popularity' command blocks into one compact example table or a single sentence noting that any keyword works with `npx clawhub search`, keeping only the `--sort` flags as a separate note.
Trim or remove the generic 'Installation Tips' and 'Best Practices' sections (e.g., 'Read documentation', 'Test thoroughly', 'Check for updates regularly') — this is guidance Claude already follows and adds no skill-specific value.
Add a short numbered workflow (e.g., 1. Search ClawHub first with `npx clawhub search "<keyword>"`; 2. If rate-limited or no results, check the OpenClaw Directory and LobeHub websites; 3. Report top candidates with install command) so the fallback order across the five sources is explicit.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly lists and commands, but roughly a third of it is padding: three near-identical `npx clawhub search "..."` blocks repeated with different keywords, plus generic advice in "Installation Tips" ("Check for updates regularly") and "Best Practices" ("Read documentation", "Test thoroughly") that Claude already knows. This fits 'mostly efficient but includes some unnecessary explanation or could be tightened' — more than the 'minor instances' required for a 4. | 3 / 5 |
Actionability | Mostly executable guidance: copy-paste-ready commands like `npx clawhub search "keyword"`, `npx clawhub browse`, and `npx clawhub search --sort installs`, plus concrete URLs. Not 5 because the GitHub strategy is only a hint ("Search: `openclaw skill`") with no executable command or URL, and there is no guidance on interpreting search output or which source to prefer first. | 4 / 5 |
Workflow Clarity | As a simple single-purpose skill the core action is unambiguous (run a clawhub search, browse categories), and the Troubleshooting section gives a fallback sequence for rate limits ("Wait 1 hour", "Use alternative sources (websites)", "Search manually on GitHub"). Not 5 because the main flow is never explicitly sequenced — the body is organized by topic (sources, strategies, tips) rather than as a first-try-then-fallback workflow, and there are no verification checkpoints. | 4 / 5 |
Progressive Disclosure | The body has good structure with clear, consistent section headers (Sources, Search Strategies, Categories, Troubleshooting) and no nested or buried references — there are no bundle files and no external file references at all. Not 5 because at ~135 lines, some catalog material (the skill category listings and troubleshooting) is inlined where a one-level-deep reference file would keep SKILL.md leaner, a minor organization gap. | 4 / 5 |
Total | 15 / 20 Passed |