Content
86%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.
A well-crafted, actionable reference: executable examples, a complete parameter table, clear modes, and useful failure signaling (cold_start, ranking debug). The only weakness is mild redundancy — the cache-warming prerequisite is restated in three separate sections. Reference links to sibling skills are clean and one level deep.
Suggestions
Consolidate the cache-warming guidance: merge 'Anti-Patterns' bullet 1 and the 'When NOT to use' fallback note into the 'Build the Cache First' section to say 'crawl first' once instead of three times.
Add a short feedback loop for weak results, e.g. 'if cold_start fires, either crawl the seed domain (see wigolo-crawl) or retry with mode: "web-expansion"' — this would turn the cold_start signal into an explicit recovery step.
Tighten the 'How It Works' step 5 by trimming the long inline ranking_debug explanation into one sentence plus the include_ranking_debug table row that already documents it.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Efficient overall — a parameters table, JSON quick reference, and terse mode list assume Claude's competence with no concept-explaining padding. The "crawl first" message repeats across "Important: Build the Cache First", "Anti-Patterns", and "When NOT to use", which could be trimmed. Not 5 because of that redundancy; not 3 because the padding is minor and localized. | 4 / 5 |
Actionability | Fully executable: copy-paste-ready JSON examples covering URL, concept, domain-scoping, and cache-only cases; a complete parameter table with types, defaults, and when-to-use; named modes; a named env var for the cold-start threshold; and a worked two-step workflow. Specific examples cover the common cases — matches anchor 5. | 5 / 5 |
Workflow Clarity | "How It Works" gives a clear 5-step sequence and the cache-building workflow is concrete (crawl with sitemap strategy, then find_similar), with `cold_start` acting as a signal checkpoint to relay. Not 5 because explicit validate-then-act feedback loops are absent (e.g., what to do when ranking sources disagree beyond enabling debug). Not 3 because the sequence is clear and checkpoints are mostly present; the destructive/batch cap does not apply to this read-only query skill. | 4 / 5 |
Progressive Disclosure | No bundle files exist, and none are needed at this size: the ~85-line body is a well-organized overview (Quick Reference, Parameters, Modes, Anti-Patterns, See Also) with clearly signaled one-level references to sibling skills (../wigolo-crawl/SKILL.md, ../wigolo-search/SKILL.md). Content is appropriately inline for a skill this size; nothing that should be split out. | 5 / 5 |
Total | 18 / 20 Passed |