City Weather Poster Generator

Problem Addressed

In creating travel check-in posters, city cultural promotions, or daily weather posters, designers often need to manually integrate city landmarks, real-time weather, and artistic styles. This process involves consolidating multi-source data, visual composition, and image generation, which is time-consuming and requires professional skills, making it difficult to respond quickly to dynamic demands.

Core Workflow

The skill automates generation through six interconnected capabilities:
- City Disambiguation: Processes user input city or attraction names with automatic disambiguation; suggests using the "province + city" format for better accuracy.
- Real-time Weather Query: Fetches current weather conditions and temperature ranges via ShowAPI to ensure data timeliness.
- City Landmark Extraction: Identifies iconic buildings or attractions of the city as core elements for the poster.
- Artistic Composition Design: Designs the layout based on the selected artistic style (e.g., default ink painting) and aspect ratio (default 9:16).
- Professional Prompt Generation: Synthesizes the above elements into detailed prompts suitable for image generation.
- Poster Image Generation: Calls scripts to asynchronously execute poster generation.

The workflow strictly follows asynchronous execution:
1. Collect user requirements, including mandatory city name and optional artistic style and aspect ratio.
2. Asynchronously submit the generation task, returning only a task_id to avoid blocking.
3. Poll the task status using task_id, providing real-time feedback of progress and core data (e.g., weather information) to the user in each round until completion.
4. Parse results, extract poster image URL, standard city name, weather conditions, etc., and generate an HTML page to showcase the integrated effect.

Applicability and Notes

• ShowAPI AppKey credential showapi_appkey must be configured; otherwise, the skill prompts for acquisition first; steps include registering an account, accessing the console, and copying the AppKey.
• All tasks are forced to execute asynchronously; synchronous mode is not supported, making it suitable for backend processing scenarios.
• During polling queries, each time data is retrieved from the interface, the main content must be immediately fed back to the user to maintain interaction transparency.
• Poster image generation takes a relatively long time; be patient to poll until the task reaches completion status.
• When inputting city names, the system automatically disambiguates, but providing a more precise format is recommended to reduce ambiguity; multiple artistic styles are supported, set via the style parameter (e.g., "Ghibli manga").