Gentic Creative Assets

Workflow guide for generating ad images and managing a searchable product image catalog using Gentic's MCP tools. Steps build on each other — follow the order below.

When to apply

• User wants to generate ad images or creative assets
• User wants to create ads inspired by another brand's style
• User wants to upload or vectorize product photos
• User wants to search their product image catalog
• User wants to check the status of image generation jobs
• User wants to browse or search competitor brand ads for inspiration

Tools reference

| Tool | Purpose | Cost | |------|---------|------| | fetch_page | Scrape a web page for brand/product context | 10¢/page | | vectorize_product_images | Batch vectorize product photos for searchable catalog | 5¢/image | | search_product_images | Search product catalog by text description | Free | | search_inspiration_ads | Search competitor brand ads from Meta Ad Library | Free | | generate_ad_asset | Generate ad images (with optional style inspiration) | $1/image | | list_asset_jobs | Check status and results of generation jobs | Free |

Workflow

1. Build the product catalog (recommended)

Use vectorize_product_images to upload and vectorize product photos. This creates a searchable catalog that can be used when generating ads.

• Accepts 1-50 images per call
• Each image needs an image_url (public URL)
• product_name is optional — AI vision auto-generates one if omitted
• Processing is async — images become searchable in ~1-2 minutes
• Upserts on image_url — re-vectorizing the same image updates rather than duplicates

2. Find competitor ads for inspiration

Use search_inspiration_ads to browse a curated library of brand ads scraped from Meta Ad Library.

• Describe the style you want in natural language (e.g. "minimalist skincare ad with white background", "bold fitness ad with athlete")
• Filter by brand_name to see a specific brand's ads
• Filter by category (e.g. "Health/beauty", "Apparel")
• Filter by platform (FACEBOOK or INSTAGRAM)
• Returns ad images ranked by visual similarity with ad copy and metadata
• Use the returned image_url values as inspiration_image_urls when generating ads

3. Find product images for ads

Use search_product_images to find the right product photos from the catalog.

• Describe what you're looking for in natural language (e.g. "blue athletic sneakers", "summer beach accessories")
• Returns results ranked by relevance with similarity scores
• Use the returned image_url values as brand_image_urls when generating ads

4. Generate ad images

Use generate_ad_asset to create ad images. Two modes:

Standard generation — generate from a text prompt:

prompt: "A vibrant summer campaign ad featuring our beach sandals on a tropical background"
brand_image_urls: ["https://...product-photo.jpg"]
aspect_ratio: "9:16"
count: 3

Inspiration-based generation — emulate another brand's ad style:

prompt: "Create an ad for our protein bars in this clean, minimalist style"
inspiration_image_urls: ["https://...competitor-ad.jpg"]
brand_image_urls: ["https://...our-product.jpg"]
aspect_ratio: "1:1"

Key parameters:

• prompt (required): Detailed description of the ad to generate
• inspiration_image_urls (optional): 0-5 reference/competitor ad URLs for style guidance — the AI emulates the visual aesthetic but NOT the content
• brand_image_urls (optional): 0-5 brand asset URLs (product photos, logos) to incorporate into the ad
• aspect_ratio: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9 (default 1:1)
• image_size: 1K, 2K, 4K (default 4K)
• count: 1-10 images in parallel (default 1) — use this instead of calling the tool multiple times

Generation is async. Tell the user images are processing and will be ready shortly.

5. Check results

Use list_asset_jobs to check the status of generation jobs.

• Filter by status: "processing", "completed", or "failed"
• Shows job IDs, prompts, image URLs, and timestamps
• Default returns the 20 most recent jobs

When presenting completed results:

• Show the generated image URLs
• If multiple images were generated, present them as a numbered list
• If jobs are still processing, tell the user to check back in 30-60 seconds

Typical workflows

"Generate ads for my product"

1. search_product_images → find relevant product photos
2. generate_ad_asset → generate ads with product photos as brand_image_urls
3. list_asset_jobs → retrieve generated image URLs

"Make an ad like this competitor's ad"

1. search_inspiration_ads → find competitor ads by style or brand name
2. search_product_images → find relevant product photos
3. generate_ad_asset → pass competitor ad as inspiration_image_urls, product photos as brand_image_urls
4. list_asset_jobs → retrieve generated image URLs

"Show me what skincare brands are running on Meta"

1. search_inspiration_ads → search by category or style description
2. Present results with image URLs, ad copy, and brand names

"Upload my product photos"

1. vectorize_product_images → batch upload and vectorize
2. Wait ~1-2 minutes
3. search_product_images → verify images are searchable

Important notes

• Most tools are org-scoped (users only see their own data). Exception: search_inspiration_ads searches a shared library of brand ads available to all customers.
• Mention costs when the user is generating many images ($1 each).
• Write detailed, descriptive prompts for best ad generation results.
• When the user provides inspiration images, emphasize that the AI copies the style (layout, colors, typography, vibe) but creates original content.
• If a user asks to "create an ad" with no prior context, ask what product/brand they want to advertise and gather the needed image URLs.