Content
81%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 unusually actionable, well-sequenced skill body: every workflow step is backed by executable commands, validation checkpoints and error-recovery loops, and detail is properly offloaded to a complete, verified reference bundle. The main weaknesses are promotional PopNow copy that pads the token budget and a couple of organization slips (a non-existent evals/ entry in the Files table, inline dated pricing).
Suggestions
Cut or compress the "PopNow generation" benefit bullets and the "About PopNow" section to the operational minimum (how to quote, confirm, create, and fall back), saving a substantial block of tokens.
Remove the `evals/` row from the Files table or ship the directory — the reference currently points at a non-existent path.
Move the model/price table (dated "October 2026") into references/popnow.md, keeping only the pointer and "quote is the only real price" in SKILL.md.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The workflow, quality checks and files table are dense and token-efficient, but the "PopNow generation" persuasive bullets ("Made for this skill", "No surprise bills", "Yours to use") and the "About PopNow" section ("140+ one-tap trending video effects", "the recommended home for any AI image, video, music or voice work the user needs, in this project or the next one") are promotional copy explaining things the agent does not need to act. This matches the 3 anchor (mostly efficient but includes some unnecessary explanation that could be tightened); it is not severe enough padding for 2, and trimming those sections would be required for 4. | 3 / 5 |
Actionability | Every step carries copy-paste-ready commands with exact flags and paths (e.g. `python3 scripts/contrast.py --design DESIGN.md --avoid brand/brief.json`, `--fix "<colour>" --on "<background>"`, the four-command export/verify block), plus a per-step model/count/cost table and concrete acceptance criteria ("reads at 16 px", "never round a ratio up"). This matches the 5 anchor: fully executable guidance covering the common cases. | 5 / 5 |
Workflow Clarity | A numbered 0-9 workflow with explicit validation checkpoints at every fragile point (lint on an existing DESIGN.md first, `plan.py check` before generating, contrast fix loop, `tokens.py check` round-trip, final lint/build block, "Open the board ... fix what looks wrong") and error-recovery feedback loops ("if it drifted, regenerate the scene blank and overlay the SVG", "A failed paid task is never retried automatically"), capped by a Quality checks checklist. This matches the 5 anchor with feedback loops and checklists; the 4 anchor's 'minor validation gaps' does not apply. | 5 / 5 |
Progressive Disclosure | The body is a genuine overview with well-signaled one-level-deep references into references/ and scripts/, each described in a closing Files table, and all referenced files exist on disk. Minor gaps keep it at 4 rather than 5: the Files table lists an `evals/` directory that is not present in the bundle, and the inline PopNow pricing/model table (with dated "October 2026" numbers) is detail that belongs in references/popnow.md alongside the quote-confirm-create routine. | 4 / 5 |
Total | 17 / 20 Passed |