Content
88%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 lean, well-sequenced instruction skill: the process is explicitly staged with validation checkpoints, an iteration loop, and an exit checklist, and the prose assumes Claude's competence with no padding. The main gaps are the absence of a worked example demonstrating the visual-scenario format it mandates, and a stray unexplained "STARTER_CHARACTER = 🎨" line at the top of the body.
Suggestions
Add one short worked example under "Show Scenarios Visually" (e.g., a before/after config snippet or input → output pair) so the skill itself demonstrates the show-don't-tell format it prescribes.
Remove or explain the stray "STARTER_CHARACTER = 🎨" line — as written it is unexplained noise that breaks the markdown structure of the body.
Optionally show a concrete sample of the options-then-decide interaction (a two-option tradeoff laid out with the follow-up question) to make step 5 copy-paste ready in conversational form.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean throughout ("Stay in design mode. Resist jumping to implementation.", "Visual beats prose.", "Then: implementation.") with zero padding and no explanation of concepts Claude already knows. Every section carries instruction weight, matching 'every token earns its place'; the only near-blemish is the single stray "STARTER_CHARACTER = 🎨" line, which is too small to drop it to 4. | 5 / 5 |
Actionability | As an instruction-only skill it gives mostly executable behavioral guidance: "show before and after", "Option A does X, Option B does Y. Which direction?", "Wait for input before proceeding", "POC for risky assumptions (API timing, format parsing)". It stops short of 5 because the core "Show, Don't Tell" section tells rather than shows — the Good example is described abstractly ("Show the actual input and expected output side by side") with no worked visual scenario demonstrating the format it prescribes. | 4 / 5 |
Workflow Clarity | The sequence is explicit (Problem → Research → Timeline → Scenarios → Decisions → Validation with numbered sections 1-7), an iteration feedback loop is shown in the diagram ("iterate freely" arrow back to Problem), step 6 is an explicit validation checkpoint ("POC for risky assumptions", "Visual test cases"), and "When to Exit Design Mode" is a completion checklist. This matches the anchor for clear sequence plus explicit validation plus feedback loop plus checklist. | 5 / 5 |
Progressive Disclosure | Sections are well-organized with clear headers and the skill is appropriately self-contained (no references exist or are needed, and nothing is inlined that belongs in a separate file). It does not reach 5 because the body opens with a stray non-markdown line "STARTER_CHARACTER = 🎨" — an unexplained constant that is a minor organization gap — and at ~80 lines it exceeds the under-50-line guideline that would let a self-contained skill score 5 on organization alone. | 4 / 5 |
Total | 18 / 20 Passed |