Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A comprehensive, actionable docs skill with strong templates and verification loops, weakened mainly by repetition of prohibitions and a monolithic structure with no reference files split out.
Suggestions
Consolidate the repeated prohibitions (no inline H1, don't infer paths from folder layout, docs.yml is read-only) into the single "Forbidden and discouraged" section and reference it once elsewhere instead of restating each 2–3 times.
Move the Fern MDX component catalog and/or changelog routing details into a references/ file (e.g. COMPONENTS.md, CHANGELOG_ROUTING.md) and link to them from SKILL.md to improve progressive disclosure.
Trim the "Style and voice" page-shape guidance to a tighter bulleted reference, keeping only what is not already implied by the v1 anchor files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Mostly efficient with no concept-padding Claude already knows, but key prohibitions are repeated three times ("no inline # H1", "don't infer paths from folder layout", "docs.yml is read-only") and could be tightened. | 2 / 3 |
Actionability | Provides copy-paste-ready frontmatter, routing YAML, MDX component examples, and verification commands — all concrete and executable. | 3 / 3 |
Workflow Clarity | "Where new pages live" and "Local verification" give clearly sequenced steps with explicit validation checkpoints (confirm frontmatter renders, components resolve, links work, images load). | 3 / 3 |
Progressive Disclosure | Well-organized into clear sections, but everything lives in one monolithic SKILL.md with no split-out reference bundle files; the component catalog and changelog details could be separated. | 2 / 3 |
Total | 10 / 12 Passed |