Promo Creator · skill publik Open Skill
Promo Creator
Create engaging promotional content for products, services, brands, videos, books, events, and social media pages. Use strong hooks, clear benefits, attractive visuals, concise copy, and effective calls to action.
Promo
2 0 5
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4179755/2026-10-02/6aj45sd4.jpg
# Promo Creator Create engaging, ready-to-use promotional content. The goal is to make the promotion eye-catching, clear, modern, persuasive without being pushy, easy to understand, suitable for social media, and visually appealing. ## When to use - The user wants promotional copy, a hook/headline, a description, a call to action, and/or a visual direction for anything being promoted (product, service, brand, event, channel, page, book, video, digital content). - The user wants an actual promotional image/poster/graphic generated, or a short promo video. - The user wants to revise or redo a single part of a promo already made with this skill (rewrite the copy, change the visual direction, regenerate the poster, etc.). ## Brand memory (remember the user's brand across sessions) This skill remembers the user's brand voice, style, and preferences so future promos stay consistent without re-asking. - **Before starting**, call `memory_query(query="<the user's brand, business, or promo style>")` to recall any saved brand voice, tone, colors, typography, audience, or platform preferences. Use what you find to shape the copy and visual direction; don't invent brand details if nothing is found. - **After finishing a promo**, call `memory_store` to save durable brand facts worth remembering across sessions, e.g. brand name, tone of voice, preferred colors/typography, target audience, and go-to platforms. Use `kind='preference'` for style/tone/visual preferences and `kind='fact'` for stable brand facts (e.g. brand name, product). Use a dedicated `namespace='promo-creator'` to keep this skill's brand memory organized. - **When the user corrects or updates a brand detail** (e.g. "actually our tone is more playful"), call `memory_query` to find the old memory's `id`, then `memory_forget(memory_id=...)` to remove it, and `memory_store` the corrected version. - Do not store one-off task progress, temporary drafts, or sensitive credentials. ## Promotional approach (always follow) Focus on: - What is being promoted. - Why people should care. - The main benefit or value. - What makes it interesting or different. - A clear call to action. Avoid exaggerated claims, fake urgency, and misleading information. Keep wording concise and easy to read, especially for mobile screens. Make the first few words or the main visual strong enough to capture attention quickly. ## Steps 1. **Recall brand memory**: call `memory_query` (see Brand memory above) to pull any saved brand voice/style/preferences. Then clarify the brief (if not already given): what is being promoted, who is the target audience, which platform(s) it's for, and any brand/style preferences. If the user uploaded a brand logo or reference image, note it for use in the visual step. 2. **Write the promotional copy** (deliver as text in your reply): - A strong headline. - An attention-grabbing hook (first line). - A short, benefit-focused description. - What makes it interesting or different. - A clear call to action that fits the content (e.g. Watch now, Learn more, Follow for more, Read the story, Try it today, Share with a friend, Discover more). Use only a CTA that fits. 3. **Write the visual direction** (deliver as text in your reply): describe imagery, typography, composition, and colors that match the brand and audience. Keep the main message visually prominent and avoid overcrowding. 4. **Save the copy + visual direction** as a document: call `create_document(content=<full promo copy + visual direction in Markdown>, name="Promo — <short title>", entity_type="promo_copy")`. If the user later asks to tweak it, use `update_document(asset_id=..., content=...)` on the same document. 5. **Generate the promotional image/poster** (if the user wants a visual, or by default when they ask for a poster/graphic): - Call `generate_image` with `model='t2i'` (text-to-image) when there is no reference image, or `model='i2i'` when the user provided a brand logo/reference image