crun-url-promo-generator

Generate promotional marketing images (posters, banners, ad graphics, social media posts) or promotional video ads (product showcase videos, brand commercial clips, ad reels) directly from a website U

SKILL.md

GitHub

Crun URL Promotional Media Generator

Use this skill to extract key selling points, product highlights, and brand aesthetics from a webpage URL (e.g., e-commerce product link, SaaS landing page, store homepage, article link) and transform them into professional AI promotional images or promotional videos using Crun image/video models.

Use ../../../runtime/crun_cli.py, ../../../catalog/models.json, and the shared child skills:

  • ../../crun-model-router/SKILL.md — model routing & live schema inspection
  • ../../crun-account-credits/SKILL.md — balance & affordability estimation
  • ../../crun-task-runner/SKILL.md — task creation, monitoring, recovery, result delivery (the authority)

Trigger

Activate for any request involving: generating promotional images, posters, ad graphics, or video ads from a URL or web link — e.g., crun-url-promo-generator, website link to poster, product link to promo video, URL to marketing image, web page to video ad, link to promotional clip — even if the user never mentions "Crun" or a specific model name.


Execution Workflow

[ User Request ] (URL + Optional Additional Notes/Preferences)
    │
    ├── 1. URL Analysis & Content Extraction
    │       ├─ Extract page title, product name, core selling points, key visual elements & color tone
    │       └─ Extract product image/logo (if accessible) → Upload: python <runtime>/crun_cli.py upload <local-file>
    │
    ├── 2. Media Type & Style Resolution
    │       ├─ Resolve Output Type: Promotional Image (Poster / Graphic) vs Promotional Video (Commercial / Ad Clip)
    │       ├─ Resolve Visual Style & Aspect Ratio (Interactive AskUserQuestion if unspecified)
    │       └─ Synthesize structured visual & narrative prompts (English)
    │
    ├── 3. Model Routing & Schema Inspection
    │       ├─ Pick image model (T2I/I2I) or video model (T2V/I2V) based on media type & reference availability
    │       └─ Inspect live schema: python <runtime>/crun_cli.py models describe --model <model>
    │
    ├── 4. Estimate & Confirmation Gate
    │       ├─ Estimate credits (crun-account-credits): task estimate --model <model> --input-file <input.json>
    │       └─ Present extracted selling points & task summary → Require explicit user OK before spending
    │
    └── 5. Task Creation & Result Delivery (crun-task-runner)
            ├─ Create task: python <runtime>/crun_cli.py task create --model <model> --input-file <input.json>
            ├─ Poll status: python <runtime>/crun_cli.py task wait --task-id <task_id>
            └─ Deliver local media path & inline image/video preview

Step 1 — URL Analysis & Content Extraction

  1. Extract Webpage Information: Use web browsing or search capabilities (read_url_content, search_web) to read the target URL's content.

    • Product / Brand Name: Identify the primary subject or service name.
    • Core Selling Points: Extract 2–4 key features, benefits, or value propositions (e.g., "AI-powered", "Ultra lightweight", "24/7 battery life", "Premium organic ingredients").
    • Visual Tone & Theme: Identify the visual style of the brand/product (e.g. minimalist high-tech, luxury elegance, vibrant energy, eco-friendly natural).
    • Product Imagery / Logo: If product key visual images or logos are present, fetch/upload them via crun_cli.py upload <file> to obtain a Crun file_url for image-edit (I2I) or image-to-video (I2V) workflows.
  2. Fallback for Inaccessible URLs: If fetching the URL directly fails (e.g., anti-scraping or paywall), ask the user for a quick text snippet, screenshot, or product image, and proceed seamlessly without halting the workflow.


Step 2 — Media Type, Style & Format Resolution

If the user did not specify the output media type, aspect ratio, or style in their request, present choices via interactive option buttons (AskUserQuestion):

  1. Media Type:

    • Promotional Image (Poster / Graphic): Static ad poster, product display graphic, social media banner, feature infographic.
    • Promotional Video (Commercial / Ad Clip): 5–10s dynamic product spotlight video, commercial clip, social media story ad.
  2. Visual Style:

    • Modern Tech / Minimalist: Clean geometric shapes, sleek lighting, soft metallic/matte textures.
    • High-End Luxury / Premium: Studio lighting, deep rich contrast, gold/dark slate accents, high-fashion aesthetic.
    • Vibrant & Dynamic: Bold color contrasts, energetic composition, modern pop/gradient visual accents.
    • Cinematic Commercial: Dramatic key lighting, shallow depth of field, realistic environment lighting.
  3. Aspect Ratio:

    • 3:4 / 9:16: Vertical (Redbook, Instagram Reels, TikTok, Mobile Ad).
    • 16:9 / 3:2: Horizontal (Website Hero Banner, YouTube, Desktop Display).
    • 1:1: Square (E-commerce main picture, Social feeds).

Step 3 — Model Selection & Prompt Generation

Model Routing Matrix

Select preferred models based on output media type and whether source media was extracted/uploaded:

Output Type Input Source Preferred Models (Text-Driven) Preferred Models (Reference-Driven, supports_reference: true) Recommended Aspect
Promotional Image URL Content Only bytedance/seedream-5-pro, openai/gpt-image-2-premium, qwen-image-3.0-pro bytedance/seedream-5-pro, qwen-image-edit-2.0-pro, google/nano-banana-pro 3:4, 16:9, 1:1
Promotional Video URL Content Only kling/kling-v2-master, minimax/video-01-live, vidu/vidu-v2 kling/kling-v2-master, bytedance/omni-human-pro, kling/kling-v1-6-pro 9:16, 16:9

Prompt Structure (English)

Image and video models perform best when visual context is specified in structured English. Translate extracted selling points and visual requirements into clear visual prompts:

  • Promotional Image Prompt Template: A high-end commercial promotional poster for [PRODUCT_NAME]. Highlighting key features: [SELLING_POINTS]. Set in a [VISUAL_STYLE] environment with professional studio lighting, crisp product textures, modern typography composition space, and elegant visual atmosphere. Masterpiece quality, 8k resolution.

  • Promotional Video Prompt Template: A cinematic promotional commercial video for [PRODUCT_NAME]. Smooth dynamic camera motion showcasing [SELLING_POINTS]. [VISUAL_STYLE] visual aesthetic, vibrant studio lighting, realistic reflections, professional advertising cinematography, fluid motion.

After picking the model, inspect its live schema before constructing payload:

python <root>/runtime/crun_cli.py models describe --model <model>

Step 4 — Estimate & Confirmation Gate

Read ../../crun-account-credits/SKILL.md. Estimate the exact input payload before creating any task:

python <root>/runtime/crun_cli.py task estimate --model <model> --input-file <input.json>

Require affordable: true. Present a clear confirmation summary card containing:

  • Extracted Product / Brand: [PRODUCT_NAME]
  • Key Selling Points Identified: [SELLING_POINTS]
  • Target Output: [Media Type] ([Aspect Ratio])
  • Selected Model: [MODEL_NAME]
  • Estimated Credit Cost: [CREDITS]

Wait for explicit user confirmation via interactive buttons before spending credits.


Step 5 — Create & Deliver

Read ../../crun-task-runner/SKILL.md for task lifecycle authority. Create once and poll:

python <root>/runtime/crun_cli.py task create --model <model> --input-file <input.json>
python <root>/runtime/crun_cli.py task wait --task-id <task_id> --timeout-seconds 120

Follow crun-task-runner for all timeout recovery and status checks.


Delivery Format

Deliver the completion summary with normalized metadata and media result:

✅ Promotional media generation completed!
- Task ID: <task_id>
- Source URL: <url>
- Media Type: <Image / Video>
- Model: <model>
- Credits Spent: <credits>
- File Path: <local_media_path>

[Inline Preview of Image / Video]