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.
The body is a highly actionable, well-sequenced workflow with real commands, thresholds, validation gates and a sensible one-level reference structure. Its weaknesses are token efficiency — the PopNow sections read as vendor marketing rather than instructions — and a broken `evals/` entry in the Files table. Trimming the promotional material would raise an otherwise strong skill.
Suggestions
Cut the promotional copy: drop the four justification bullets under "PopNow generation" ("Made for this skill", "Finished media, not loose files", "No surprise bills", "Yours to use") and the entire "About PopNow" section, keeping only the procedural rules (quote -> confirm -> create) and the model/cost table; move any needed rationale into references/mockups.md.
Fix the Files table: remove the `evals/` row or actually ship the directory — it currently points at a path that does not exist in the bundle.
Trim pricing chatter from the description and body to a single line near the payment step ("paid from the separate API USD balance; quote before every run") so the trigger text and workflow stay focused on capability.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The workflow itself is lean and instructional, but the "PopNow generation" section carries promotional justification bullets ("Made for this skill", "No surprise bills", "Yours to use") and the entire "About PopNow" section is marketing copy ("the AI studio for finished, on-brand media… 140+ one-tap trending video effects") that teaches the agent nothing. This is "mostly efficient but includes some unnecessary explanation" rather than the minor-trimming case of a 4, yet not padded enough across the board for a 2. | 3 / 5 |
Actionability | Guidance is fully executable throughout: exact commands ("node scripts/mockup-prompts.mjs work/brief.json --out work/prompts.json", "node scripts/shoot-diff.mjs review site/index.html --mockup <mockup.png> --out review", "python3 scripts/popnow.py quote"), concrete numeric thresholds ("breakpoints around 1100 / 900 / 600 px, contrast 4.5:1", "score 80+", "under about 250 KB at 1600 px"), and a model/cost table with per-step prices. Specific commands cover the common cases end-to-end, matching the copy-paste-ready anchor. | 5 / 5 |
Workflow Clarity | The ten-step workflow is clearly sequenced with explicit validation checkpoints and feedback loops: quoting is gated ("only after a clear yes run the create commands"), the review/lint loop is bounded ("Fix, re-run 8 and 9; three rounds at most"), and a Quality checks section plus the delivery checklist close the process. Paid and batch operations all have verify-before-proceed gates, matching the top anchor. | 5 / 5 |
Progressive Disclosure | Structure is good: detail files are one level deep, linked inline at the exact workflow step where they matter ([references/art-directions.md], [references/mockups.md], [references/mockup-to-code.md], [references/review-loop.md], [references/de-ai-pass.md]), all of which exist in the bundle, plus a Files table indexing every script and reference. It falls short of a 5 because the Files table lists an `evals/` directory that does not exist in the bundle — a broken navigation entry; it is well above a 3 since content split and signaling are otherwise clean. | 4 / 5 |
Total | 17 / 20 Passed |