WaClaw Tutorial and Practical Test - Agent: One-Click Generation of a Full Set of E-commerce Main Images

I uploaded a flat lay image, and five minutes later, the main image for Taobao, the white background image for Amazon, the product recommendation image for Xiaohongshu, the long image for the product details page, and the model video all appeared. It only cost me 9.9 yuan.

For those of you in e-commerce, who hasn't tried a few things? AI Raw image tools?

I've used quite a few, but the conventional ones... AI Images generated by the tool AI The image quality is very poor, resulting in stiff and unnatural images that are vastly different from the platform's image style.

In e-commerce, every image, from the main image to the product details page, serves a specific purpose: to showcase the product's overall appearance, highlight its selling points, and generate a desire to purchase. Different platforms have different guidelines regarding size, composition, and content.

Want to AI Producing consistently high-quality e-commerce images that can be delivered requires more than just the ability to create raw images; it also demands an understanding of design, products, and platform rules.

Most importantly, you need to understand the store's style; otherwise, you'll inevitably have to make manual adjustments in the end.

Today, let's take a hands-on test of WaClaw, the new e-commerce platform for lobsters launched by Huiwa.

We open the WaClaw application, select WaClaw, and simply tell WaClaw what type we need. WaClaw will...automaticBreak down the task and produce a finished product that conforms to the platform's rules.

Basic product main image

I first uploaded a flat-lay photo of the sun-protective clothing, and then directly told WaClaw:

Prompt wordsGenerate a high-quality main image for this purple sun-protective clothing item in the picture, suitable for listing on Taobao.

Here's a little tip: For more consistent image quality, it's best to...automaticSelect the model and turn it off.For images, manually select Pic 3.1; for videos, select Video 3.0.。

IPrompt wordsThe color of the clothes was mistakenly written as purple. (WaClaw) automaticI identified the product reference image and corrected the color description.Prompt wordsWaClaw also recognized all the design details that weren't mentioned at all, such as the hidden zippered pockets, dropped shoulders, and the horizontal yoke dividing line on the back.

WaClaw first confirmed that the product belonged to the apparel category and was intended for use as a main image on Taobao. Then, it searched for relevant regulations, clarifying my unclear understanding...Prompt wordsThe instructions are translated into specific image requirements, such as using a portrait image with a 3:4 aspect ratio.

Finally, WaClaw willautomaticOnly after completing a platform self-check to confirm that the image ratio, subject display, and text watermark all meet the platform's requirements will the results be delivered.

The entire task consumed 15 points, and the generated main image looks like this:

The overall quality is excellent, with clear images, natural lighting, and an outdoor setting that's perfect for showcasing sun-protective clothing. The consistency between the product and the people in the images is also consistent; the color tone and style of all six images are unified, conveying a high-quality brand image.。

WaClaw also generated product attribute images and color selection images based on the additional selling points I provided, ensuring that the e-commerce main image displayed all the necessary information.

With each seasonal update, we can launch 30 new SKUs, and we no longer need to pay a large sum of money for photography studios, models, and post-production.

The same request was sent to others. AI Try the tools:

The samePrompt words,generally AI The generated images were quite ordinary and didn't really fit the platform's style.

Let's try generating another set of Amazon main images.

Prompt wordsGenerate another set of Amazon main images.

WaClaw will reorganize the image creation plan according to the platform rules and regenerate the main image at a 1:1 ratio according to the requirements of Amazon's apparel category, making the switching between different platforms very smooth.

This used 24 points. Let's take a look at the final result:

The characters were changed to European and American models, the text was changed to English, and the overall visual effect was changed to the Amazon style.

WaClaw canautomaticIntegrate platform rules into raw image requirements to generate matching materials.。

To sell the same product from Taobao to Amazon, you don't need to find a new team to plan, shoot, and produce English-language materials. You only need to add one request.fastWe obtained a set of initial product images that conform to the browsing habits of overseas consumers.

Multi-pose model pictures

Model photos are also a high-frequency requirement for e-commerce images, especially in the clothing category. Real-life photos of models can directly affect consumers' first impression of the product and even affect product conversion.

However, it's not enough to just find any good-looking model to wear the clothes in a model photo.

If a brand has a fixed visual style, or if we want to refer to the models, scenes, and lighting of a store, we can first send the reference materials to WaClaw for study, and then generate model images.

We open the file in the list on the left.

You can copy the Taobao link; WaClaw will...automaticIdentify all images in the link, including the main image, detail images, etc.

In the dialogue, we let WaClaw learn the style first, and then generate images. A maximum of 10 materials can be selected as references each time.

Prompt wordsThe style of the model photos in my store is as follows: @Image 1 @Image 2 @Image 3 @Image 4 @Image 5 @Image 6 @Image 7 @Image 8 @Image 9 @Image 10. Please study the details of this set of photos and the overall style of this set of photos, including the model, accessories, lighting, background, composition, etc. The subsequent clothes and models should be used to generate model photos with a consistent style for the store.

WaClaw will fully understand the store's style and product characteristics, and generate model images that are consistent with our brand's overall visual style.

This task costs 30 points, and the composition and lighting are similar to the style of the lesson.

Previously used AI To maintain brand consistency in graphic design, constant fine-tuning and experimentation with different approaches are necessary.Prompt wordsNow we will provide reference images directly. AI The learning and operational barriers have been significantly lowered.

WaClaw also has a verypracticalSmall features.

If you are satisfied with the generated result, you can click directly in the upper right corner of the dialog box.Save as Skill。

WaClaw will document the entire process of this conversation, including platform guidelines, model style, image proportions, product display methods, and the details we adjusted; these will all become parameters that can be reused in the future.。

In the future, when there are similar products, we can simply select this skill and upload a new image to generate similar materials in batches.

Generate product detail page

A product detail page not only needs to be aesthetically pleasing, but also needs to contain structured information, understand user needs, and showcase the key points that users care about.

We can also directly learn from the layout of the product detail pages of best-selling products:

Prompt words@Image 3 is my model, @Image 1 and @Image 2 are my products, and @Image 4 is the layout (collage) of the product details page that I want to reference. Please help me break it down into multiple single images and provide a generation plan. Note the generation ratio of each image. If you need a combination of images and text, you also need to plan carefully.

WaClaw analyzes the layout structure of the reference diagram and plans it according to different blocks. Before the actual generation, it will provide a planning outline and confirm the key parameters with us.

All generated materials supportOne-clickBatch download:

We can also directly use WaClaw to stitch together and generate a single long image for the entire details page.

This task cost a total of 120 points.

Previously, creating a product detail page required finding reference templates, retouching images one by one, and going through several rounds of revisions. A single product detail page could take a graphic designer half a day to complete. During peak periods such as bulk new product launches and major sales events, this process could easily become a bottleneck for the entire team's productivity.。

Now WaClaw can directly learn from excellent templates and create a solution in just a few minutes, significantly improving efficiency.。

There's no need to worry about production capacity limitations when creating product detail pages in bulk.

The same approach is also suitable for creating Amazon A+ charts:

Off-site product recommendation materials

Many merchants use AI The images they created look like high-end but cold advertisements at first glance, making it difficult for them to gain exposure on social media.

WaClaw can distinguish the differences in visual language between commercial displays on e-commerce platforms and casual sharing on social media platforms, generating images that match the platform's tone by adjusting lighting, camera angles, and composition.。

Prompt wordsUsing the pink sun-protective clothing (@Image 1 & @Image 2), generate a set of 4 images showcasing the garment in multiple ways. All 4 images use the same tag (@Image 3) and are set in the same indoor scene (white walls, wood-grain floor, white window on the left, bright and soft natural light). The overall style is reminiscent of Xiaohongshu fashion blogger product recommendation images—gentle, fresh, and with bright, harmonious colors.

This round of tasks consumed 72 points. Let's take a look at the results:

The scenes and styles are consistently excellent, and the overall effect is very similar to a professional selfie, relaxed and natural.

In the past, if we wanted to collaborate with influencers, we had to contact them, coordinate schedules, send samples, and handle post-production.

Now, by simply providing product images and reference styles, WaClaw can generate a batch of influencer marketing materials in minutes, enabling influencer campaigns and social media content creation, while also providing a lower-cost trial-and-error approach.

Mashup Video

Besides still images, video content is also an indispensable part of operations. Fast-paced, mashup videos are more suitable for social media platforms and short video channels.

Prompt wordsA 9:16 vertical outdoor cycling video was generated: a short-haired girl wearing a pink hooded sun-protective jacket and white shorts rides a white bicycle through a summery, tree-lined park. The camera transitions from a full-body shot to medium and close-ups and detailed shots, showcasing the adjustable hood drawstring, smooth zipper, lightweight and breathable fabric, embroidery on the chest, and adjustable hem, before concluding with a wide shot of her cycling from behind. The overall effect is fresh and natural, resembling a real-life mobile phone shot, with smooth camera movements, accompanied by upbeat music and simple Chinese subtitles.

This task costs 150 points. Let's take a look at the final result:

The full-body displays and close-up details of the models are incredibly realistic, and the overall pacing and transitions between shots are very natural. For clothing items, this kind of dynamic presentation is more persuasive than static images.

Product recommendation video

Fashion videos that bloggers on social media platforms are making can also be generated directly.

Prompt wordsPlease generate a vertical, girly dress-up video based on the four outfit photos and model photos I uploaded: the same model with the same sweet makeup, scenes rotating, each scene showing the model standing or sitting with a side view and smiling, the overall effect is bright, high-saturation, Korean film-style girly feel.

This task costs 120 points. Let's take a look at the generated results:

The characters and costumes maintain a good consistency, andautomaticWith background music, a well-structured outfit video sample can be created in just a few minutes.

Normally, hiring a model to shoot a fashion video would cost several hundred yuan, but with WaClaw, it only takes 120 points, less than one-tenth the cost of a real shoot.

One person plus WaClaw, 9.9 yuan to complete all the images, text and videos for a store's entire season's new product launch.。

Use different AI When using tools to create e-commerce content, you need to clearly list out all the details and key points, just like training a newcomer. The quality of the finished product is still very random and may not match the style of your store.

WaClaw, through its self-developed models and scenario scheduling, is like a seasoned, knowledgeable employee. It not only understands the compliance and hit-product logic of different platforms such as Taobao, Amazon, and Xiaohongshu, but also identifies the brand's visual DNA.

What's most amazing is that WaClaw also allows you to learn and accumulate experience.

We can save the highlights of best-selling cases and successful processes as Skills through dialogue, turning them into reusable experiences.

In the past, maintaining a consistent visual style was a capability typically reserved for large brands. WaClaw learns from stores' color schemes, compositions, and copywriting habits, allowing ordinary people to achieve brand-level visual productivity at a low cost.

A strategic and aesthetically pleasing individual paired with an e-commerce expert. AI AgentThis can unleash the productive energy of an entire medium-sized team.

This forces e-commerce practitioners to break free from tedious mechanical execution and focus their energy on irreplaceable core decision-making aspects such as user insights, product selection planning, and traffic management.

Original link:The biggest fear when creating e-commerce images AI The taste was solved by WaClaw.