Person Landmark Check-In Photo Generator

Problem Solved

When users have only a person reference image but want a realistic landmark check-in photo, manual prompting often misses environment, lighting, pose, and identity consistency. This skill targets travel check-in photos, virtual trip mementos, and social media visuals. It turns a person image, location, and optional pose into an asynchronous image generation task and keeps reporting task state, instead of issuing one opaque long-running request.

How It Works

• Intent parsing: reads refer_image, location_prompt, and optional pose requirements, then extracts scene type and structured intent.
• Prompt optimization: builds an optimized_prompt covering environment, person, lighting, and style, and uses negative_prompt to exclude cartoon-like, blurry, or other non-photorealistic elements.
• Async submission: calls the ShowAPI generation task and returns only a task_id without blocking.
• Status polling: polls until the task reaches SUCCEEDED, then exposes core fields such as check_in_photo_url.

Boundaries and Notes

Real calls require showapi_appkey; without credentials, the skill performs a one-shot simulation preview and does not run the real async task. The person photo can be a standard Base64 string or a public URL, image ratio defaults to 3:4, and pose is optional. All real tasks are asynchronous, making this suitable for workflows that can wait for generated results and display the image, but not for lightweight calls that require an immediate synchronous response.