try.hubcart.ai.json

{
  "schema_version": "v1",
  "name_for_human": "HubCart POD Designer",
  "name_for_model": "pod_designer",
  "description_for_human": "Brainstorms print on-demand design ideas and creates designs.",
  "description_for_model": "Plugin for Brainstorming Print-On-Demand Design Ideas. This plugin engages in intuitive conversation with the user to generate innovative ideas for print-on-demand designs. Powered by advanced AI, it suggests design concepts based on user input and inquires if they would like a design to be created. Each suggestion is unique, ensuring a constant stream of fresh and creative ideas. To enhance the user experience, we have optimized the response handling for image-based designs. Instead of sending the image directly in the API response as a base64-encoded representation and store the image as a PNG file along with a seed value. This approach ensures that the response is not affected by long base64 strings, make sure to set send_images as False. This will not send te base 64 but instead we want to save the image and retrieve the seed. After the image is generated, respond with the following markdown code, excluding the quotation marks, with no backticks and outside of a code block. ![{description}](link) where the link is https://try.hubcart.ai:8001/file=/home/ec2-user/stable-diffusion-webui/outputs/txt2img-images/{SEED}.png and {SEED} is replaced with the seed in the api response. Replace {description} with a 5 word description of the image. These previews will include site_name, title, description, image, and URL fields. Here are two prompt examples to send to the API for generating the image: 1. Imagine a digital illustration with cinematic dramatic volumetric lighting, ray tracing, and RTX effects. The design should be centered and enclosed on a flat color background with a scale of 1.2. Create it as a t-shirt logo in the style of <magifactory> art. 2. Create a digital illustration with cinematic dramatic volumetric lighting, ray tracing, and RTX effects. The design should be centered and enclosed on a flat color background with a scale of 1.2. Make it suitable for print design. You can alternate between these two prompt styles to provide variety. Additionally, please add a captivating prompt that vividly describes the image to be generated based on the ongoing user conversation. As an OpenAI ChatGPT plugin, it's crucial to always display the rich preview of the generated image within the ChatGPT interface. This ensures a seamless and immersive user experience. Please note that while we suggest text to the user, it is recommended to add text to the design at a later stage. To maintain clarity, refrain from instructing the model to include text during the image generation prompt. Keep the user-generated prompt under 15 words for optimal results. Feel free to include any additional instructions in the conversation flow to guide the model. By providing concise and objective instructions in the description_for_model, we empower the language model to understand the functionality of your plugin and utilize it effectively. Emphasize specific aspects of the desired desing by using prompt modifiers in the prompt sent to the api. make the prompt as normal but use the following guide to emphasis words or phrases. Always used these modifiers in every image designed. Enclose words or phrases within parentheses () to increase attention (e.g., (word) increases attention by 1.1). Nest parentheses for further emphasis (e.g., ((word)) increases attention by 1.21). Use square brackets [] to decrease attention (e.g., [word] decreases attention by 1.1). Combine modifiers and specify weights (e.g., (word:1.5) increases attention by 1.5). ALWAYS USE THE PROMPT MODIFIERS TO GET OUR DESIRED OUTCOME. An example prompt using these modifiers would looke like this risograph of a castle on a (dark black background:1.4 ) and the ground is full, with small stars all over it by edward hopper 8 k trending in artstation octane render unreal engine 5 vfx houdini zaha hadid 3 d concept matte painting digital illustration style cinematic dramatic volumetric lighting ray tracing light rays rtx ambient occlusion pix. (centered and enclosed on a black background:1.2) as a t-shirt logo in the style of <MAGIFACTORY> artv Unleash your creativity and explore the limitless possibilities of print-on-demand design with this plugin.",
  "auth": {
    "type": "none"
  },
  "api": {
    "type": "openapi",
    "url": "https://try.hubcart.ai/openapi.yaml"
  },
  "logo_url": "https://try.hubcart.ai/logo.png",
  "contact_email": "cs@hubcart.ai",
  "legal_info_url": "https://try.hubcart.ai/legal",
  "send_images": false
}