Content
90%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 body is a lean, highly actionable API guide: executable examples for every core operation, explicit read-before/verify-after validation with error recovery, and tight, well-labeled sections. The only gaps are the prose-only standalone-file creation path lacking a verification step and a single-file structure that inlines reference-style detail.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Statements like "A native Figma spec is GitHub-Flavored Markdown (GFM) under a `SPEC_BLOCK` node" convey only Figma-specific facts Claude would not know; there is no padding and no explanation of Markdown, GFM, or Figma basics. Not 4 because no section contains over-explanation that could be trimmed. | 5 / 5 |
Actionability | Copy-paste-ready calls such as `await figma.spec.createOnCanvasAsync({...})`, `figma.spec.getContent({ specNodeId, offset: 100, maxLines: 50 })`, `figma.spec.editContent({ edits: [{ oldText, newText }] })`, and `figma.spec.comments.create({ anchorText, contextText, message })` cover the common create/read/edit/comment cases with exact parameters and post-write error checks. Not 4 because the shown examples have no gaps. | 5 / 5 |
Workflow Clarity | The main edit loop has explicit checkpoints — "Use `getContent` with the same `specNodeId` before edits and to verify writes" plus `throw new Error('...was not applied')` verification — and recovery guidance like "Add surrounding context to make it unique", so the destructive-operation cap does not apply. It is not 5 because the standalone-file creation path ("call `whoami`, choose a destination plan with `specs_enabled: true`, and call `create_new_file`...") is prose-only with no verification checkpoint — a minor validation gap consistent with anchor 4. | 4 / 5 |
Progressive Disclosure | The single file is well-organized into clear sections (Discover, Create, Content methods, Critical rules, Style, Images, Comments) with no nested or buried references. It is not 5 because at ~110 lines with no reference split it exceeds the under-50-line simple-skill case, and reference-like detail (Style, Images) is inlined; not 3 because nothing is buried and navigation is easy. | 4 / 5 |
Total | 18 / 20 Passed |