Generates AI images via Alibaba DashScope APIs with async task polling and local download.
npx skills add https://github.com/openakita/openakita --skill tongyi-image
Please help me install this Agent Skill. Skill: tongyi-image Source: https://github.com/openakita/openakita/tree/main/plugins/tongyi-image Command: npx skills add https://github.com/openakita/openakita --skill tongyi-image
Creating production-ready images for posters, product listings, and creative concepts normally requires design tools or manual work across multiple AI services. This Skill turns text prompts or reference images into finished pictures through Alibaba Cloud DashScope, handling the async task lifecycle, polling, and local file download automatically.
Ask the assistant to generate an image with a prompt like "a girl playing guqin in a bamboo forest, ink wash style" using the tongyi_image_create tool after configuring your DashScope API key in Settings.
High-intent search queries and answers about installing and using this skill.
Call the tongyi_image_create tool with a prompt, model (default wan27-pro), and size such as 1024*1024. The tool submits an async DashScope task, polls until completion, and returns image URLs, local paths, and asset IDs.
It supports seven modes: text-to-image, image editing, style repaint, background generation, outpainting, sketch-to-image, and e-commerce suites. Each mode maps to specific DashScope models like wan2.7-image-pro or wanx-background-generation.
Yes, a DashScope API key from the Alibaba Cloud Bailian console is required. Configure it in the plugin Settings under dashscope_api_key, or set the DASHSCOPE_API_KEY environment variable before creating tasks.
Common causes are a missing API key, which returns a configuration error, or DashScope content moderation rejecting sensitive prompts, which fails without retry. Modify the prompt wording and resubmit the task.
Yes, generated images are published to the Asset Bus with asset_ids that downstream tools like seedance-video can consume directly. The storyboard plugin can also export shots as batch request bodies for bulk image generation.
Avoid it for video generation, which belongs to the seedance-video skill, and for precise mask-based local retouching, which the image-edit plugin handles better. Large e-commerce batches also increase cost, so test with one image first.
hawkongzConvert images into Chinese text descriptions using Alibaba Cloud Bailian vision models.
ToqsickGenerate motion posters from product images and brand assets.
SpaceZephyrTurn a story idea into a structured storyboard plan with consistent panel images.
prime-skillsRemove or replace unwanted image regions using masks or descriptions.
