Content
68%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 well-structured, concrete, and information-dense across all three sync operations, with proper use of an external schema reference. Its main weaknesses are the lack of validation checkpoints around the batch PUSH operation and references to helper functions that have no implementation in the bundle.
Suggestions
Add explicit validation checkpoints to the PUSH workflow: run DIFF first to preview pending changes, and verify the Figma API response (variable counts, assigned IDs) before writing the lock file — this is a batch write to a shared design file and currently has no validation gate.
Either ship implementations for `parseCSSTokens()`, `classifyToken()`, and `hexToFigmaColor()` in a scripts/ directory (with paths in the body) or explicitly state Claude must implement them from the described rules — they are currently named as if they already exist.
Remove the Quick Reference metadata table ("Complexity: High | 4 projects") at the top of the body; it duplicates frontmatter information and adds tokens without operational value.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and operational (API endpoints, config examples, type-mapping tables) with almost no explanation of concepts Claude already knows. However, the Quick Reference table ("Complexity | High | 4 projects") duplicates frontmatter metadata and the four-tier parsed-token JSON example could be trimmed, so it fits anchor 4's 'minor instances that could be trimmed' rather than anchor 5's 'every token earns its place'. | 4 / 5 |
Actionability | Guidance is mostly executable: concrete API endpoints with auth headers, request-body structure, rate-limit policy, naming conventions, lock-file format, and DIFF exit codes. It falls short of anchor 5 because functions like `parseCSSTokens(cssString)`, `classifyToken(name, value)`, and `hexToFigmaColor()` are referenced as if they exist, yet no scripts/ directory or implementation is provided in the bundle. | 4 / 5 |
Workflow Clarity | Each of the three operations has a clearly numbered, well-sequenced step list, but PUSH is a batch write to a shared Figma file with no pre-push validation checkpoint (no dry-run or DIFF-first preview) and no post-push verification beyond writing the lock file after 'a successful push'. Per the rubric's batch-operation cap, missing validation gates cap this at anchor 3 even though the sequences themselves are clear. | 3 / 5 |
Progressive Disclosure | The config schema is appropriately split into a real, clearly signaled one-level-deep reference (`references/figma-config-schema.md`) with only a minimal example inlined. It is not anchor 5 because the body is a fairly monolithic full instruction set — detail like the complete lock-file format and troubleshooting matrix lives inline with only one external reference, fitting anchor 4's 'most content appropriately placed; minor organization gaps'. | 4 / 5 |
Total | 15 / 20 Passed |