Content
62%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, highly actionable skill with excellent workflow clarity including explicit blocking gates and validation checkpoints. However, it suffers significantly from verbosity — the main SKILL.md tries to be both an overview and a detailed reference simultaneously, resulting in a very long document that doesn't efficiently use the token budget. The progressive disclosure structure exists (good reference links) but is undermined by too much inline detail.
Suggestions
Move the lengthy 'Image Generation Tools' resolution logic (backend selection, fallback chains, SVG prohibition) to a reference file like references/image-backend-resolution.md, keeping only a 2-3 line summary in SKILL.md.
Move 'Batch Generation Policy' and 'Confirmation Policy' details into references/workflow.md where they naturally belong, retaining only one-line summaries in the main file.
Consolidate the 'Changing Preferences' section and EXTEND.md path priority table into the existing references/config/ directory, since these are configuration details not needed on every invocation.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~300+ lines. It over-explains tool resolution logic (Image Generation Tools section alone is massive), batch policies, confirmation policies, and preference management in exhaustive detail. Much of this (e.g., how to resolve backends across multiple runtimes, detailed EXTEND.md path priority tables) could be in reference files. Claude doesn't need lengthy explanations of fallback chains inline. | 1 / 3 |
Actionability | The skill provides highly concrete, actionable guidance: specific file naming patterns (NN-{type}-{slug}.md), exact directory structures, explicit YAML frontmatter fields, concrete command flags (--type, --style, --palette, --preset), clear table-driven options for each step, and specific tool invocation patterns. The workflow steps are executable and specific. | 3 / 3 |
Workflow Clarity | The 6-step workflow is clearly sequenced with explicit blocking gates (Step 1.5 is marked ⛔ BLOCKING, Step 3 is a hard gate before Steps 4+, Step 5 has a BLOCKING requirement for prompt files before generation). Validation checkpoints, retry policies (retry failed items once), and error recovery (text correction via new prompt file) are all explicit. The confirmation policy provides clear skip conditions. | 3 / 3 |
Progressive Disclosure | The skill does reference multiple external files (references/workflow.md, references/styles.md, references/prompt-construction.md, etc.) with clear navigation links, which is good structure. However, the SKILL.md itself is bloated with content that should live in those reference files — the Image Generation Tools section, Batch Generation Policy, and Confirmation Policy are all lengthy inline sections that duplicate or preview what the references contain. The main file should be leaner with more content pushed to references. | 2 / 3 |
Total | 9 / 12 Passed |