Ask Generative AI Model to Generate a Style
Goal
Generate a custom style for image requests using a Generative AI model.
Constraints:
- Console log all processes
- Request sent, including the request body
- Response received, including the response body
- Write file output, including the file name and path
- The output must be a valid JSON object, suitable for direct use in image generation APIs.
- The style object must include exactly the fields and structure as specified in the latest sample request/response in the input file:
ai-labs/recraft/generate-image-style-recraft.md
. This input file is the SINGLE SOURCE OF TRUTH for required fields and structure. Do not hardcode or assume fields—always read from the input file. - All errors, edge cases, and exceptions must be logged to the console with clear error messages.
- The script must validate the response from the AI model before writing output. If invalid, log the error and do not write to file.
- If the output file already exists, append a timestamp to the filename to avoid overwriting.
- The script must include comprehensive comments explaining each major logic block.
Inputs
Model API Specifics:
- See:
ai-labs/recraft/generate-image-style-recraft.md
for model-specific instructions, sample payloads, and authentication details. - Use the provided API endpoint, authentication method, and payload structure as specified in that file.
Example Request Body
- The request and response structure must always match the latest sample in
ai-labs/recraft/generate-image-style-recraft.md
. - Do not use or reference static examples here—always refer to the input file for the canonical structure.
Outputs
- Write the resulting style object to:
ai-labs/recraft/styles-recraft.json
- If the file exists, create a new file with a timestamp suffix (e.g.,
styles-recraft-2025-04-14T20-49-42.json
). - The output structure must match the canonical structure in
ai-labs/recraft/generate-image-style-recraft.md
.
Step-by-Step Workflow
- Read model API instructions and sample payloads from
generate-image-style-recraft.md
. - Construct a prompt and request body as shown in the input file.
- Send the request to the specified AI model API.
- Log the outgoing request, including the full request body.
- Log the incoming response, including the full response body.
- Validate the response for required fields and correct format, based on the input file.
- Write the validated style object to
styles-recraft.json
(or timestamped variant). - Log the file output, including the file name and path.
- Log and handle all errors, including invalid responses and file write issues.
Edge Cases
- If the AI model returns an invalid or empty response, log an error and do not write output.
- If the output file already exists, do not overwrite—create a new file with a timestamp.
- If authentication fails, log the error and exit.
- If any required field is missing in the response, log the error and do not write output.
Documentation
- The script must include comprehensive comments at each major block, explaining the logic and any model-specific considerations.
Related Files
- Model API instructions & canonical sample:
ai-labs/recraft/generate-image-style-recraft.md
- Output file:
ai-labs/recraft/styles-recraft.json
Example Usage
shell
python ai-labs/recraft/generate-style-recraft.py
This prompt is intended to ensure robust, transparent, and reproducible style generation workflows for image APIs. Always use the input file as the single source of truth. ;)