SuperAPI AI Directory

wechat-sticker-assets-designer

Design & creation

View repository

Updated 10-15 21:56

About this skill

WeChat Sticker Assets Designer (微信表情包配套素材设计师)

This skill is intended to help the Agent play the role of a professional designer, analyze the user's existing sticker pack IP image, and strictly follow the specifications of the WeChat Sticker Open Platform to sequentially generateBanner、Appreciation guidance imageandAppreciation thank-you image。


🎨 Core Constraints (Global Constraints)

  1. IP Character Consistency (Character Consistency):
    • When generating all 3 images, the core IP features in the user's reference image (for example: hairstyle, accessories, facial proportions, classic color palette, etc.) must be strictly retained and inherited. It must be ensured that it is the same character portrayed in different scenes.
  2. Emoji expressions are strictly prohibited:
    • In all generated images,absolutely must notany Emoji symbols appear (such as 😂, ❤️, 👍, etc.).
  3. Image format specifications:
    • When constructing the drawing prompt, clearly describe style words that help present a PNG texture, such as clean, flat, and vector illustration, to ensure clean edges.
  4. Sequential Output (Sequential Output):
    • The workflow must be followed, generating and displaying one image at a time by independently calling the tool,strictly prohibitedto combine and stitch multiple images into one long image.

📏 Image Generation Specifications (Detailed Specs)

First image: WeChat Banner

  • Size ratio:750x400 pixels (width-to-length ratio about 1.875:1, set the ratio in the prompt, such as --ar 15:8 or 1.875:1 aspect ratio)。
  • Image content:
    • Design a complete scene with this IP character as the core, withstorytellingandrich details(such as: the character exploring in space, camping in the forest, or creating at a desk, etc.).
    • The color tone is lively and bright.
    • Background constraints (extremely important): The backgroundmust not use white, andmust not use a transparent background, it must have a full colored background, so as to form a clear visual boundary in WeChat's light-colored interface.
  • Text constraints:Any text or garbled characters are strictly prohibited in the image。

Second image: Appreciation guidance image (Solicitation Image)

  • Size ratio:750x560 pixels (width-to-length ratio about 1.34:1, set the ratio in the prompt, such as --ar 4:3 or 1.34:1 aspect ratio)。
  • Image content:
    • The character needs to show an expression and action of "looking forward to support", "cute and endearing", or "holding a bowl with both hands / begging for a tip", to enhance the appeal of appreciation.
    • Background constraints:It is strictly prohibited to use white or near-white light-colored backgrounds and edges. The background must be solid-colored and darker than the WeChat page background color.
  • Copy requirements:
    • Chinese text must be perfectly integrated in a suitable position in the image: "您的赞赏是我们最大的动力!".
    • The font, color, and style of the text must be highly consistent with the overall illustration style of the image.

Third Image: Acknowledgment Image

  • Size Ratio:750x750 pixels (1:1 square ratio, set the ratio in the Prompt --ar 1:1)。
  • Image Content:
    • The character shows a happy state of "extremely grateful," "celebrating," "cheering and throwing confetti," or "making a heart shape with both hands in thanks."
    • The image has rich and saturated colors, creating a festive and joyful interactive atmosphere that inspires sharing.
    • Background Constraints:Strictly prohibited from using white or near-white light-colored backgrounds and edges。
  • Copy Requirements:
    • Chinese text must be perfectly integrated into a suitable position in the image: "Thank you for your generous appreciation!".
    • The typography and design of the text must be harmonious with the illustration style of the image.

🛠️ Usage Guide

1. Quick Start

No need to manually install dependencies; just run the script directly. The tool will automatically create a virtual environment (.venv) and install the required dependencies.

# 横幅 Banner (目标 750x400,采用居中 center 裁剪)
./skills/wechat-sticker-assets-designer/scripts/run.sh "PATH_TO_BANNER" --width 750 --height 400 --anchor center

# 赞赏引导图 (目标 750x560,采用靠上 top 裁剪以防切掉顶部文字)
./skills/wechat-sticker-assets-designer/scripts/run.sh "PATH_TO_SOLICITATION" --width 750 --height 560 --anchor top

# 赞赏致谢图 (目标 750x750,采用居中 center 缩放)
./skills/wechat-sticker-assets-designer/scripts/run.sh "PATH_TO_ACKNOWLEDGMENT" --width 750 --height 750 --anchor center

2. (Optional) Manual Installation

If you prefer to manage the environment manually:

python3 -m venv .venv
source .venv/bin/activate
pip install -r skills/wechat-sticker-assets-designer/requirements.txt
python3 skills/wechat-sticker-assets-designer/scripts/crop_and_resize.py ...

🛠️ Execution Workflow

After the Agent activates this skill, it must strictly follow the four steps below:

  1. Step 1: Analyze the IP Character Style。
    • Read and analyze the [expression example image] or character description provided by the user, extract and summarize the character's line strokes, color palette, and signature features (such as "a Nezha character with double hair buns and a red ribbon").
  2. Step 2: Generate the Banner。
    • Construct a drawing Prompt describing a rich story scene that meets the Banner specifications, and call generate_image to generate.
    • Check the generated image: ensurethere is no text at allandthere is no white/transparent background。
  3. Step 3: Generate the Appreciation Guide Image。
    • Construct a Prompt describing the character's cute, expectant pose, and add typography instructions containing the copy "Your appreciation is our greatest motivation!", and call generate_image to generate.
    • Check whether the background is non-white.
  4. Step 4: Generate the Appreciation Thank-You Image。
    • Construct a Prompt describing the character's happy pose of celebrating and thanking with a heart gesture, and add typography instructions containing the copy "Thank you for your generous appreciation!", and call generate_image to generate.
  5. Step 5: Output Result Verification and Cropping (CRITICAL CHECK & CROP)。
    • Important Note (Conditional Cropping by Script): Because the drawing tool may output a 1:1 image by default, the Agent mustuse the built-in cropping script of this skill after generation to perform conditional size verification (cropping and scaling are only actually performed when the size does not match):
      • Sample script command to run (recommended: use the run.sh script):
        # 横幅 Banner (目标 750x400,采用居中 center 裁剪)
        ./skills/wechat-sticker-assets-designer/scripts/run.sh "PATH_TO_BANNER" --width 750 --height 400 --anchor center
        
        # 赞赏引导图 (目标 750x560,采用靠上 top 裁剪以防切掉顶部文字)
        ./skills/wechat-sticker-assets-designer/scripts/run.sh "PATH_TO_SOLICITATION" --width 750 --height 560 --anchor top
        
        # 赞赏致谢图 (目标 750x750,采用居中 center 缩放)
        ./skills/wechat-sticker-assets-designer/scripts/run.sh "PATH_TO_ACKNOWLEDGMENT" --width 750 --height 750 --anchor center
        
    • Before delivery, the Agent must explicitly confirm that the final pixel dimensions of the three assets fully meet the requirements。
    • Hard Constraint: If any image is missing due to any drawing tool failure,the image generation tool must be immediately called again to complete the generation, never deliver only part of the materials.
    • Finally, copy/save these 3 independent images to the output directory specified by the user (if not specified, create by default the output/sticker_assets_[theme] directory), and report the absolute paths, frame specifications, and previews to the user.