Content
92%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 skill body: concise, concrete, and self-contained, with a tool map, a verification-driven workflow with a feedback loop, and genuinely Figma-specific mappings. The only nit is a couple of mapping rows that restate CSS knowledge Claude already has under a "non-obvious" heading.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and dense — a tool table, a 6-step workflow, and hard acceptance thresholds with almost no padding. A few mapping-table rows restate knowledge Claude already has ("Fixed | width: Npx", flex-direction equivalence), which is minor over-explanation that could be trimmed, matching the anchor-4 description rather than a fully token-perfect 5. | 4 / 5 |
Actionability | Guidance is fully executable for an MCP-tool skill: exact tool names (figma/get_metadata, figma/get_variable_defs), the required input shape (Figma URL including node-id), concrete output targets (src/styles/tokens.css), a data-testid convention, and numeric acceptance criteria ("spacing within 4px, token colors exact, font family and weight exact"). | 5 / 5 |
Workflow Clarity | The 6-step workflow has a clear sequence, an explicit verification step (get_screenshot side-by-side plus DOM bounding-box comparison with thresholds), and a feedback loop ("Outside threshold → fix the token mapping or ask design; re-run from step 2"), matching the top anchor. | 5 / 5 |
Progressive Disclosure | The skill is under 50 lines with no need for external reference files; content is organized into clear sections (MCP tools, Workflow, Figma→CSS mappings), and its only pointers (the frontend-design skill and the docs URL) are clearly signaled one level deep — the simple-skill case the scoring notes describe. | 5 / 5 |
Total | 19 / 20 Passed |