Cosa contiene

Un file SKILL.md (la guida operativa) + un modello di partenza. Un'anteprima:

Product Poster

When to use

• The user is launching a product or running a promo and needs a hero image for a product page or a social post.
• The user is hosting an event (course launch, offline salon, livestream announcement) and needs a shareable poster.
• The user hands over a long product write-up and needs it distilled into "one graphic that explains the selling points, the price, and how to buy".
• The user is running feed ads and needs the first frame to grab attention in 1 second (headline + visual hook).
• The user already has a poster draft and wants it reviewed against layout and conversion elements.

Workflow

1. Break down the selling points (Input: product brief/event plan → Action: distill → Output: selling-point list)
   Distill 1 core selling point (≤ 10 words, answering "why you") + 2–3 supporting points (≤ 10 words each). Method: delete every adjective in the brief — the facts left standing are your selling points; if facts run short, ask the user. Never pad with "ultimate" or "premium".
2. Set the price anchor (Input: pricing → Action: three-layer presentation → Output: price plan)
   Present price in three layers: original price (struck through, anchors perceived value) → current price (big type) → urgency (limited time/quantity, e.g. "first 100 only"). No real original price? Then no strikethrough — a fake original price is price fraud; for free events, make "FREE" the big price.
3. Write the call to action (Input: user goal → Action: write CTA → Output: CTA copy)
   CTA ≤ 8 words, a direct action: "Buy now", "Scan to register", "Book a slot" — never a passive "Learn more". Pair it with one urgency element (pick one): a deadline ("ends tonight at midnight") or scarcity ("37 seats left" — the number must be real).
4. Pick a layout (Input: selling points + price + CTA → Action: choose one of three → Output: layout plan)
   Layout A — centered hero: big product/person image centered, selling points around it, price bar at the bottom; best for a single hero product. Layout B — split columns: image left, copy right (or reverse), selling points stacked vertically; best for multi-point products. Layout C — three bands: top = hook headline, middle = product + points, bottom = price + CTA; best for events/courses. After choosing, mark safe margins (keep price and CTA 8–10% clear of every edge).
5. Write the image prompt (Input: layout + points + price + CTA → Action: fill the template.md skeleton → Output: complete prompt)
   Write the prompt in English and include: canvas ratio (e-commerce 3:4 / social 1:1 / landscape 16:9), the layout, primary color hex, product description. All display text — headline, price digits, CTA — must be marked "leave blank for text overlay" and added later; never let the model render words (one wrong digit in a price is a disaster).
6. Generate and self-check (Input: prompt → Action: generate + verify → Output: final or revision list)
   Check five things: ① the 1-second test — squint at the image and say what it sells, for how much, how to buy; ② price digits match the user's numbers exactly; ③ CTA inside safe margins and big enough; ④ no garbled text, no stray elements; ⑤ still legible as a phone thumbnail. Fail any of these → go back to step 5 and revise the prompt.

Anteprima del modello

1. Selling points (fill in before designing)

Core selling point (exactly 1, ≤ 10 words)

Supporting points (2–3, ≤ 10 words each)

Price anchor (three layers — numbers must be real)

Il pacchetto della skill (SKILL.md + modello) è scritto in inglese.

Acquista il pacchetto completo della skill — $4.99 Pro — tutte le 511 skill